# forck.live — 24–30 August 2026 > AI lab announcements, news reports and podcasts. forck.live follows what companies announce about their own products. Press may corroborate an official story; it never leads and never appears as an item of its own. A missing date stays missing. A title is the publisher's words — where a company does not publish in English, an English rendering is listed separately as `Title in English` and named as this site's translation, never as theirs. This file is one issue of the weekly archive: every confirmed first-party announcement of ISO week 2026-W35, 24–30 August 2026, newest first. The window never moves, so this URL always names the same seven days — but a post published inside them and detected later is added when it arrives, so a closed week can still gain entries. Every entry is one first-party announcement, and carries the same eight fields in this order before any others: Company · Announced · Category · Coverage · Announcement · Group · Source · Record `Source` is the company's own page the announcement was traced to and `Record` is this site's page for it. Both are safe to cite. `Coverage` is a number of outlets, or `not counted` — which means this record reached no count of its own, never that nobody wrote about it. `Category` is this tracker's own analysis, and is authoritative wherever it is stated. `not stated` means either that nothing has analysed the post yet or that the analysis placed it outside this taxonomy; neither is a verdict about the announcement. `Announcement` is whether the post is news of something the company did, rather than a how-to, a customer story, a changelog listing or a staffing note. `Group` is which of this record's groups the entry falls in, carried on the entry itself so it holds where the groups are headings and where they are not. One of: `models`, `covered`, `announcements`, `routine`. Anything after `Record` is held detail that only some announcements carry. - This week's page: https://forck.live/week/2026-W35 - The current issue, and the catalogue of the other weeks: https://forck.live/briefing.md - Catalogue: https://forck.live/llms.txt ## Announcements ### OpenClaw — OpenClaw 2.0, Accidentally - Company: OpenClaw (openclaw.ai) - Announced: 2026-08-30T12:00:00+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://openclaw.ai/blog/openclaw-2-accidentally/ - Record: https://forck.live/items/17759-openclaw-2-0-accidentally - Subject: OpenClaw OpenClaw released its largest update, OpenClaw 2.0, built by 933 contributors with over 16,000 pull requests. The update simplifies installation, rebuilds the browser app as a first-class experience, and introduces shared cloud sessions for multiplayer collaboration. It touches every part of OpenClaw including messaging, memory, skills, models, automations, and security. ### Midjourney — Edit Image Quality Update - Company: Midjourney (midjourney.com) - Announced: 2026-08-29T00:35:46+00:00 - Category: model-update - Coverage: not counted - Announcement: yes - Group: models - Source: https://updates.midjourney.com/edit-image-quality-update/ - Record: https://forck.live/items/18580-edit-image-quality-update - Subject: Midjourney V8 Midjourney updated its V8.2 edit model to improve image quality. Users who experienced issues in the previous 24 hours are encouraged to try again. ### xAI — Grok Bot now works with X - Company: xAI (x.ai) - Announced: 2026-08-29 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://x.ai/news/grok-bot-and-x - Record: https://forck.live/items/5037-grok-bot-now-works-with-x - Subject: Grok Grok Bot now integrates with X, allowing users to connect their X account, get free X API credits, and perform actions like searching posts, reading timelines, and checking mentions. ### GitHub — GitHub Copilot in Visual Studio — August update - Company: GitHub (github.com) - Announced: 2026-08-28T20:24:20+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://github.blog/changelog/2026-08-28-github-copilot-in-visual-studio-august-update-2 - Record: https://forck.live/items/4670-github-copilot-in-visual-studio-august-update - Subject: Copilot GitHub Copilot in Visual Studio added organization-level custom agents, thinking effort controls (Low, Medium, High) for supported models, improved model management with pinning and cost information, usage tracking, and a Git agent for reviewing uncommitted changes and commits before pull requests. The update is available to all GitHub Copilot plan tiers. ### GitHub — GitHub Copilot weekly releases — August 24 - Company: GitHub (github.com) - Announced: 2026-08-28T20:13:22+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://github.blog/changelog/2026-08-28-github-copilot-weekly-releases-august-24 - Record: https://forck.live/items/4734-github-copilot-weekly-releases-august-24 - Subject: Copilot GitHub Copilot weekly releases on August 24 include new features such as team sessions in Slack and Teams, a generally available Customize tab that brings MCP servers, plugins, skills, and canvases together, experimental support for WSL, ability to split and move tabs, send browser preview to external browser, default execution and permission modes, improved plugin management, session restoration, faster Copilot CLI on Rust runtime, enterprise controls for JetBrains, continuing agent sessions between apps, complementary model second opinions, a new single-pane Agents layout, detailed chat usage by model, model pinning and comparison, sharing custom agents across organizations, plan consumption management, and Git agent for reviewing changes before PRs. ### Amazon — Batch write and discover records in Amazon SageMaker Feature Store - Company: Amazon (amazon.com) - Announced: 2026-08-28T19:31:05+00:00 - Category: api-release - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://aws.amazon.com/blogs/machine-learning/batch-write-and-discover-records-in-amazon-sagemaker-feature-store/ - Record: https://forck.live/items/4638-batch-write-and-discover-records-in-amazon-sagemaker-feature-store - Subject: Bedrock / Nova Amazon announces two new APIs for SageMaker Feature Store: BatchWriteRecord for batch writing up to 25 records across multiple feature groups, and ListRecords for enumerating record identifiers within a feature group. ### Perplexity — GLM 5.3 - Company: Perplexity (perplexity.ai) - Announced: 2026-08-28T16:36:56+00:00 - Category: api-release - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://docs.perplexity.ai/docs/resources/changelog#august-2026 - Record: https://forck.live/items/5056-glm-5-3 - Subject: Sonar The Agent API and Router API now support the perplexity/glm-5.3 model with specified pricing. ### Amazon — How Decathlon runs demand forecasting at scale with Chronos-2 - Company: Amazon (amazon.com) - Announced: 2026-08-28T16:22:30+00:00 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://aws.amazon.com/blogs/machine-learning/how-decathlon-runs-demand-forecasting-at-scale-with-chronos-2/ - Record: https://forck.live/items/4635-how-decathlon-runs-demand-forecasting-at-scale-with-chronos-2 - Subject: Bedrock / Nova Decathlon, a major sporting goods retailer, selected Chronos-2 as a core component of its forecasting stack after evaluating multiple time series foundation models. In benchmarks on Decathlon's own retail data, fine-tuned Chronos-2 consistently outperformed all other evaluated models across both 12-week and 52-week horizons, while even zero-shot Chronos-2 matched or surpassed the existing production baseline. Chronos-2 also met efficiency requirements, running inference for 25,000 products in under 2 minutes per cutoff. ### Amazon — Spreading the load: How Salesforce met Multi-AZ HA with SageMaker Inference Components - Company: Amazon (amazon.com) - Announced: 2026-08-28T16:20:40+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://aws.amazon.com/blogs/machine-learning/spreading-the-load-how-salesforce-met-multi-az-ha-with-sagemaker-inference-components/ - Record: https://forck.live/items/4633-spreading-the-load-how-salesforce-met-multi-az-ha-with-sagemaker-inference - Subject: Bedrock / Nova AWS announced the SchedulingConfig parameter for the CreateInferenceComponent API in Amazon SageMaker AI Inference Components, giving customers fine-grained control over IC copy placement across instances and Availability Zones to achieve Multi-AZ high availability. The post describes how Salesforce used this capability to meet their compliance requirements. ### GitHub — Upcoming changes to GitHub Copilot policies and billing - Company: GitHub (github.com) - Announced: 2026-08-28T11:37:19+00:00 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://github.blog/changelog/2026-08-28-upcoming-changes-to-github-copilot-policies-and-billing - Record: https://forck.live/items/4585-upcoming-changes-to-github-copilot-policies-and-billing - Subject: Copilot GitHub announces upcoming changes to Copilot policies and billing: billing changes for Copilot Business/Enterprise starting September/October 2026, convergence of Copilot Chat experiences into a single unified experience no earlier than September 28, 2026, and change of default review effort level for Copilot code review from Lite to Balanced starting September 28, 2026. ### Tencent — Tencent Releases and Open-Sources Tencent Hy4 preview - Company: Tencent (tencent.com) - Announced: 2026-08-28T08:48:45+00:00 - Category: open-weight-release - Coverage: not counted - Announcement: yes - Group: models - Source: https://www.tencent.com/tencent-releases-and-open-sources-tencent-hy4-preview/ - Record: https://forck.live/items/7438-tencent-releases-and-open-sources-tencent-hy4-preview - Subject: Hunyuan Tencent released and open-sourced Hy4 preview, a 770B-parameter large language model with 49B active parameters and a context window exceeding 1M tokens. The model is available through Tencent products and via API, with free access on WorkBuddy and CodeBuddy for two weeks. API pricing is set at USD 0.834 per million input tokens, USD 2.501 per million output tokens, and USD 0.042 per million tokens for cache hits. ### OpenAI — Our decision on Cursor following its acquisition by SpaceX - Company: OpenAI (openai.com) - Announced: 2026-08-28T06:00:00+00:00 - Category: not stated - Coverage: 2 outlets - Announcement: no - Group: covered - Source: https://openai.com/index/our-decision-on-cursor-following-its-acquisition-by-spacex - Record: https://forck.live/items/5018-our-decision-on-cursor-following-its-acquisition-by-spacex - Subject: GPT / ChatGPT / API OpenAI announced it will wind down its contract providing OpenAI models to Cursor following Cursor's acquisition by SpaceX. ### OpenAI — Supporting Thailand’s next generation of AI startups - Company: OpenAI (openai.com) - Announced: 2026-08-28T02:00:00+00:00 - Category: partnership-acquisition - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://openai.com/index/supporting-next-generation-ai-startups-thailand - Record: https://forck.live/items/4574-supporting-thailand-s-next-generation-of-ai-startups - Subject: GPT / ChatGPT / API OpenAI and Thailand's MHESI launch an eight-week accelerator to help 10 health, wellness, and education startups turn AI prototypes into trusted products. ### Databricks — Fast, fault-tolerant PyTorch training on AI Runtime - Company: Databricks (databricks.com) - Announced: 2026-08-28T01:15:00+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://www.databricks.com/blog/fast-fault-tolerant-pytorch-training-ai-runtime - Record: https://forck.live/items/17495-fast-fault-tolerant-pytorch-training-on-ai-runtime - Subject: Mosaic AI Databricks explains how PyTorch's distributed asynchronous checkpointing and data pipeline design improve training goodput at scale. The post details that GPU failures are expected in large jobs, and that using distributed checkpoint (DCP) with async_save enables frequent, low-cost checkpointing, reducing recovery time and idle GPU spend. ### Apple — Agent Seer: Synthesizing Scenarios from Specification Understanding - Company: Apple (apple.com) - Announced: 2026-08-28 - Category: research-paper - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://machinelearning.apple.com/research/agent-seer-synthesizing-scenarios - Record: https://forck.live/items/10304-agent-seer-synthesizing-scenarios-from-specification-understanding - Subject: Machine Learning Research Apple researchers present Agent Seer, a pipeline that synthesizes realistic evaluation scenarios for tool-using AI agents from a single Model Context Protocol (MCP) specification without examples, live tool access, or domain-specific tuning. The pipeline generates graded scenarios and multi-turn dialogues, achieving strong tool-calling correctness and conversational coherence across seven MCP specifications. The analysis finds that parameter schema complexity is the strongest correlate of quality variation, with argument value accuracy as the dominant failure mode. ### Apple — LLMs Are Not (Consistently) Bayesian: Quantifying Internal (In)consistencies of LLMs’ Probabilistic Beliefs - Company: Apple (apple.com) - Announced: 2026-08-28 - Category: research-paper - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://machinelearning.apple.com/research/llms-not-consistently-bayesian - Record: https://forck.live/items/10303-llms-are-not-consistently-bayesian-quantifying-internal-in-consistencies-of - Subject: Machine Learning Research Apple researchers introduce a technique to study LLMs as information processing rules, using the deviation from Bayesian updates to measure internal inconsistencies in how LLMs update probabilistic beliefs from evidence. Their experiments show that non-Bayesian heuristic updates often outperform exact Bayesian updates in downstream tasks, suggesting LLMs' probabilistic world models are misspecified. The measure can provide diagnostics for issues in LLM-powered inferential systems. ### Sesame — TurnBench: measuring the timing of conversation - Company: Sesame (sesame.com) - Announced: 2026-08-28 - Category: research-paper - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://www.sesame.com/blog/turnbench - Record: https://forck.live/items/8043-turnbench-measuring-the-timing-of-conversation - Subject: Maya / CSM Sesame released TurnBench, an open-source benchmark for evaluating turn-taking dynamics in spoken dialogue. The benchmark includes a public leaderboard, interactive conversation viewer, and self-serve scoring on the dev set. It was developed in partnership with Carnegie Mellon University, Mundo AI, Oto, and researchers from several other institutions. ### Amp — No Mailmap Required - Company: Amp (ampcode.com) - Announced: 2026-08-28 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://ampcode.com/news/no-mailmap-required - Record: https://forck.live/items/7705-no-mailmap-required - Subject: Amp Amp now allows users to set a separate Git identity for commits, with options for personal default, project-specific, or workspace-wide rules. ### Amp — Amp on iOS & macOS - Company: Amp (ampcode.com) - Announced: 2026-08-28 - Category: availability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://ampcode.com/news/amp-on-ios-and-macos - Record: https://forck.live/items/7704-amp-on-ios-macos - Subject: Amp Amp released an iOS app via TestFlight and a macOS app via direct download, enabling users to control their orbs from anywhere. ### Anthropic — Claude for Teachers, now available for U.S. K-12 schools and districts - Company: Anthropic (anthropic.com) - Announced: 2026-08-28 - Category: availability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://claude.com/blog/claude-for-teachers-now-available-for-schools-and-districts - Record: https://forck.live/items/7414-claude-for-teachers-now-available-for-u-s-k-12-schools-and-districts - Subject: Claude Anthropic made Claude for Teachers available to U.S. K-12 schools and districts as a free Enterprise offering, with enterprise features and K-12 terms. Qualifying organizations can sign up by June 30, 2027 for a full year of free access. ### Anthropic — How Anthropic employees use Claude Tag - Company: Anthropic (anthropic.com) - Announced: 2026-08-28 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://claude.com/blog/how-anthropic-employees-use-claude-tag - Record: https://forck.live/items/7413-how-anthropic-employees-use-claude-tag - Subject: Claude Anthropic employees use Claude Tag to produce customer-ready collateral, compile weekly issue reports, and run legal review, all within Slack. The tool can generate a document from a Slack thread, consolidate feature requests from across channels, and compile incident reports from multiple Slack channels. ### Together AI — GLM-5.3 vs. GLM-5.3 Flash on DeepSWE: Cost, Coding, and Routing - Company: Together AI (together.ai) - Announced: 2026-08-28 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://www.together.ai/blog/glm-5-3-vs-glm-5-3-flash-on-deepswe-cost-coding-and-routing - Record: https://forck.live/items/5017-glm-5-3-vs-glm-5-3-flash-on-deepswe-cost-coding-and-routing - Subject: Inference platform An analysis comparing GLM-5.3 and its distilled sibling GLM-5.3 Flash on the DeepSWE benchmark, finding that the Flash variant retains 94% of solved tasks at one-seventeenth the cost but loses reliability rather than capability, with a cascade strategy solving 80.9% of tasks at $1.70 each. ### Hugging Face — The Open ASR Leaderboard Adds Its First Global South Language - Company: Hugging Face (huggingface.co) - Announced: 2026-08-28 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://huggingface.co/blog/open-asr-leaderboard-global-south - Record: https://forck.live/items/4657-the-open-asr-leaderboard-adds-its-first-global-south-language - Subject: Platform The Open ASR Leaderboard adds two new evaluation sets, Monsoon en-IN and Monsoon hi-IN, for Hindi and Indian English, to address bias and improve representation of Global South languages and diverse speaker demographics. ### Cohere — Generative AI for business: Use cases, benefits, and adoption - Company: Cohere (cohere.com) - Announced: 2026-08-28 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://cohere.com/blog/generative-ai-for-business - Record: https://forck.live/items/4573-generative-ai-for-business-use-cases-benefits-and-adoption - Subject: Command Generative AI for business, covering use cases, benefits, and challenges. ### Midjourney — Edit Model for V8 - Company: Midjourney (midjourney.com) - Announced: 2026-08-27T23:32:15+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://updates.midjourney.com/edit-model-for-v8/ - Record: https://forck.live/items/18581-edit-model-for-v8 - Subject: Midjourney V8 Midjourney began testing its first V8.2 image edit model, which supports editing images with instructions, generating images with up to four image references, inpainting, outpainting, personalization, moodboards, and srefs. The model is accessible via drag-and-drop, the edit button in the lightbox, the edit tab on the website, or the --edit url command on Discord. The company updated UIs on both midjourney.com and alpha.midjourney.com and is asking users to report edge cases and share feedback. ### Amazon — Build agentic creative workflows with Amazon Quick and fal - Company: Amazon (amazon.com) - Announced: 2026-08-27T23:04:22+00:00 - Category: capability-change - Coverage: not counted - Announcement: no - Group: routine - Source: https://aws.amazon.com/blogs/machine-learning/build-agentic-creative-workflows-with-amazon-quick-and-fal/ - Record: https://forck.live/items/4536-build-agentic-creative-workflows-with-amazon-quick-and-fal - Subject: Bedrock / Nova Amazon announces an integration between Amazon Quick and fal via the Model Context Protocol (MCP), enabling agentic creative workflows for media teams. The post describes how Amazon Quick serves as an agent workspace and orchestrator, fal provides generative media models, and MCP standardizes the connection. Two example workflows are outlined: creating an eight-panel storyboard and prototyping a music-video concept. The integration allows context retention, human review at creative gates, and reuse of workflow instructions as Skills. ### GitHub — Copilot code review: Resolution reasons and expanded capabilities - Company: GitHub (github.com) - Announced: 2026-08-27T22:46:04+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://github.blog/changelog/2026-08-27-copilot-code-review-resolution-reasons-and-expanded-capabilities - Record: https://forck.live/items/4535-copilot-code-review-resolution-reasons-and-expanded-capabilities - Subject: Copilot GitHub Copilot code review now supports reviewing pull requests authored by bots (including Copilot cloud agent) and very large pull requests, and adds the ability to specify a resolution reason (Addressed, Won't fix, Incorrect) for review comments. ### Meta — Closed-Loop Cooling Explained: The Plumbing Behind Meta’s AI - Company: Meta (meta.com) - Announced: 2026-08-27T21:44:08+00:00 - Category: infrastructure-release - Coverage: not counted - Announcement: no - Group: routine - Source: https://about.fb.com/news/2026/08/closed-loop-cooling-explained-the-plumbing-behind-metas-ai/ - Record: https://forck.live/items/9912-closed-loop-cooling-explained-the-plumbing-behind-meta-s-ai - Subject: Meta AI Meta describes its closed-loop liquid cooling system for AI data centers, which recirculates a water-glycol coolant to cool GPU servers efficiently. The company shares its cooling designs openly via the Open Compute Project, including the IcePack liquid-cooled network rack platform announced in 2025. Meta also uses reinforcement learning to optimize cooling infrastructure. ### fal — Introducing H3 Max by fal - Company: fal (fal.ai) - Announced: 2026-08-27T19:02:36+00:00 - Category: new-model - Coverage: not counted - Announcement: yes - Group: models - Source: https://blog.fal.ai/introducing-h3-max-by-fal/ - Record: https://forck.live/items/7991-introducing-h3-max-by-fal - Subject: fal platform fal Research released H3 Max, a post-trained version of MiniMax H3 optimized for speed. According to their evaluations, H3 Max ranks first in quality and speed compared to leading video models. ### Amazon — Introducing OpenAI models on Amazon Bedrock for in-country inferencing in India - Company: Amazon (amazon.com) - Announced: 2026-08-27T18:36:08+00:00 - Category: availability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://aws.amazon.com/blogs/machine-learning/introducing-openai-models-on-amazon-bedrock-for-in-country-inferencing-in-india/ - Record: https://forck.live/items/4532-introducing-openai-models-on-amazon-bedrock-for-in-country-inferencing-in-india - Subject: Bedrock / Nova Amazon Bedrock announces the availability of OpenAI GPT-5.6 models (Terra and Luna) in India with India geographic cross-Region inference, allowing customers to process inference requests and data within India to meet data residency requirements. Both models offer a 1-million-token context window and accept text and image input for text output. ### Amazon — Introducing India cross-Region inference for OpenAI GPT-5.6 models on Amazon Bedrock - Company: Amazon (amazon.com) - Announced: 2026-08-27T18:36:08+00:00 - Category: availability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://aws.amazon.com/blogs/machine-learning/introducing-india-cross-region-inference-for-openai-gpt-5-6-models-on-amazon-bedrock/ - Record: https://forck.live/items/4525-introducing-india-cross-region-inference-for-openai-gpt-5-6-models-on-amazon - Subject: Bedrock / Nova Amazon Bedrock now supports OpenAI GPT-5.6 models (Terra and Luna) in India with India geographic cross-Region inference, enabling data residency within India. The models offer a 1-million-token context window and accept text and image input. Cross-Region inference routes requests only within India across the Mumbai and Hyderabad Regions to meet data residency requirements. ### Google Research — Planetary prediction engine: Automating global models via Earth AI - Company: Google Research (research.google) - Announced: 2026-08-27T17:37:20+00:00 - Category: research-paper - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://research.google/blog/planetary-prediction-engine-automating-global-models-via-earth-ai/ - Record: https://forck.live/items/4522-planetary-prediction-engine-automating-global-models-via-earth-ai - Subject: Research Google Research introduces the planetary prediction engine (PPE), an experimental research capability under the Google Earth AI initiative that autonomously executes the full geospatial modeling workflow from data discovery to model training, achieving improvements across diverse prediction tasks in public health, food security, environmental risk, and socioeconomics. ### Microsoft — Your agent’s guardrails have a bypass - Company: Microsoft (microsoft.com) - Announced: 2026-08-27T17:14:46+00:00 - Category: developer-tool-release - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://commandline.microsoft.com/agent-hooks-framework-neutral-ai-governance-contract/ - Record: https://forck.live/items/19129-your-agent-s-guardrails-have-a-bypass - Subject: Foundry / Responsible AI Microsoft published Agent Hooks, an open, framework-neutral governance contract for AI agents, along with SDKs in Python, TypeScript, .NET, Rust, and Go, and a 47-scenario conformance kit. The contract defines eight interception points and three verdicts to standardize guardrail enforcement across agent frameworks. ### Perplexity — DeepSeek V4 Flash 0731 - Company: Perplexity (perplexity.ai) - Announced: 2026-08-27T17:10:21+00:00 - Category: new-model - Coverage: not counted - Announcement: yes - Group: models - Source: https://docs.perplexity.ai/docs/resources/changelog#august-2026 - Record: https://forck.live/items/5063-deepseek-v4-flash-0731 - Subject: Sonar Perplexity announced support for the new model perplexity/deepseek-v4-flash-0731 in its Agent API and Router API, describing it as a fast, efficient open reasoning model with a 1M-token context window. ### Perplexity — NVIDIA Nemotron 3.5 Lightning - Company: Perplexity (perplexity.ai) - Announced: 2026-08-27T17:10:21+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://docs.perplexity.ai/docs/resources/changelog#august-2026 - Record: https://forck.live/items/5062-nvidia-nemotron-3-5-lightning - Subject: Sonar Perplexity's Agent API and Router API now support the perplexity/nemotron-3.5-lightning-30b-a3b model, an open-weight reasoning model, with pricing details. ### Perplexity — NVIDIA Nemotron 3 Ultra - Company: Perplexity (perplexity.ai) - Announced: 2026-08-27T17:10:21+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://docs.perplexity.ai/docs/resources/changelog#august-2026 - Record: https://forck.live/items/5061-nvidia-nemotron-3-ultra - Subject: Sonar Perplexity's Agent API and Router API now support the perplexity/nemotron-3-ultra-550b-a55b model at the stated pricing. ### Microsoft — Managed PostgreSQL vs. self-hosted PostgreSQL: Key benefits and trade-offs - Company: Microsoft (microsoft.com) - Announced: 2026-08-27T17:00:00+00:00 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://azure.microsoft.com/en-us/blog/managed-postgresql-vs-self-hosted-postgresql-key-benefits-and-trade-offs/ - Record: https://forck.live/items/4562-managed-postgresql-vs-self-hosted-postgresql-key-benefits-and-trade-offs - Subject: Azure AI Microsoft publishes a comparison of managed PostgreSQL (Azure Database for PostgreSQL, Azure HorizonDB) versus self-hosted PostgreSQL, discussing trade-offs in control, operational tax, security, and engineering capacity for technical decision makers. ### Google DeepMind — Gemini Omni 1.1 Flash lets you build with more control - Company: Google DeepMind (deepmind.google) - Announced: 2026-08-27T16:11:32+00:00 - Category: model-update - Coverage: 8 outlets - Announcement: yes - Group: models - Source: https://deepmind.google/blog/gemini-omni-1-1-flash-lets-you-build-with-more-control/ - Record: https://forck.live/items/4515-gemini-omni-1-1-flash-lets-you-build-with-more-control - Subject: Gemini Google DeepMind announces Gemini Omni 1.1 Flash, a new suite of creative controls and generative video capabilities including scene extension, first and last frame interpolation, 360p draft mode, 4K upscaling, and video references. ### Amazon — Deepgram deepens Amazon SageMaker AI observability with Enhanced Metrics - Company: Amazon (amazon.com) - Announced: 2026-08-27T16:11:27+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://aws.amazon.com/blogs/machine-learning/deepgram-deepens-amazon-sagemaker-ai-observability-with-enhanced-metrics/ - Record: https://forck.live/items/4516-deepgram-deepens-amazon-sagemaker-ai-observability-with-enhanced-metrics - Subject: Bedrock / Nova Deepgram announces Enhanced Metrics and Prometheus/OpenTelemetry support for its speech-to-text and text-to-speech models deployed on Amazon SageMaker AI, providing billing and usage transparency and engine-level observability directly into customers' CloudWatch accounts. ### Amazon — Reduce ASR inference costs by 75% with NVIDIA MPS on Amazon EC2 - Company: Amazon (amazon.com) - Announced: 2026-08-27T16:05:10+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://aws.amazon.com/blogs/machine-learning/reduce-asr-inference-costs-by-75-with-nvidia-mps-on-amazon-ec2/ - Record: https://forck.live/items/4517-reduce-asr-inference-costs-by-75-with-nvidia-mps-on-amazon-ec2 - Subject: Bedrock / Nova Amazon Web Services, NVIDIA, and Heidi Health announce a solution that uses NVIDIA CUDA Multi-Process Service and NVIDIA Triton Inference Server on Amazon EC2 GPU instances to reduce GPU infrastructure requirements for ASR inference by 75 percent, from 16 instances to 4, while maintaining sub-second latency at 92.1 requests per second per GPU, enabling Heidi Health to process over 2.4 million clinical consultations per week. ### Tencent — Introducing Hy4 preview - Company: Tencent (tencent.com) - Announced: 2026-08-27T16:00:00+00:00 - Category: new-model - Coverage: not counted - Announcement: yes - Group: models - Source: https://hunyuan.tencent.com/research/hy4-preview - Record: https://forck.live/items/4575-introducing-hy4-preview - Subject: Hunyuan Tencent releases Hy4 preview, a 770B-parameter model with 49B active parameters and a 1M-token context window, open-sourced and available on Tencent Cloud and OpenRouter. ### Databricks — Enhancing Agent Retrieval with Structured Chart Extraction - Company: Databricks (databricks.com) - Announced: 2026-08-27T15:00:00+00:00 - Category: research-paper - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://www.databricks.com/blog/enhancing-agent-retrieval-structured-chart-extraction - Record: https://forck.live/items/17496-enhancing-agent-retrieval-with-structured-chart-extraction - Subject: Mosaic AI Databricks demonstrates that structured extraction from charts, using its ai_parse_document function to represent charts as structured JSON, improves both retrieval and answer quality on chart-based questions. The approach, evaluated on a chart-heavy subset of ViDoRe V3 and a synthetic Chart-RAG dataset, performs competitively with larger multimodal embedding models while using a lightweight 300M-parameter text embedding model. Adding images of the top retrieved chunks at answer time further improves answer quality by 2.6 to 4 percentage points. ### Google DeepMind — Piloting the world's first double-blind AI evaluations - Company: Google DeepMind (deepmind.google) - Announced: 2026-08-27T12:59:16+00:00 - Category: capability-change - Coverage: 4 outlets - Announcement: yes - Group: covered - Source: https://deepmind.google/blog/piloting-the-worlds-first-double-blind-ai-evaluations/ - Record: https://forck.live/items/4499-piloting-the-world-s-first-double-blind-ai-evaluations - Subject: Gemini Google DeepMind announces the world's first double-blind evaluation of a proprietary frontier AI model, using cryptographic environments to prevent benchmark contamination. The pilot tests a Gemini Flash Lite model with external partners against confidential benchmarks in a privacy-preserving environment, enhancing evaluation integrity and trust in model assessments. ### Google Antigravity — Teamwork: When AI Becomes a Research Partner - Company: Google Antigravity (antigravity.google) - Announced: 2026-08-27T10:00:00+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://antigravity.google/blog/teamwork-when-ai-becomes-a-research-partner - Record: https://forck.live/items/4539-teamwork-when-ai-becomes-a-research-partner - Subject: Antigravity Antigravity announces several updates to Teamwork, a multi-agent orchestration framework, including new patterns like Iterative Coding, Distributed Coding, Long Proof, Self-Verification, and Document Review. The updates build on Gemini 3.7 Flash and are available as /teamwork-preview on all paid plans. ### OpenAI — Better answers, broader thinking: What students gain from ChatGPT and critical-thinking training - Company: OpenAI (openai.com) - Announced: 2026-08-27T09:00:00+00:00 - Category: research-paper - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://openai.com/index/what-students-gain-from-chatgpt-critical-thinking-training - Record: https://forck.live/items/4519-better-answers-broader-thinking-what-students-gain-from-chatgpt-and-critical - Subject: GPT / ChatGPT / API OpenAI published a study involving over 1,000 students that examines the impact of ChatGPT and critical-thinking training on originality and student performance in a real-world university assignment. ### Vercel — The best workflow engine is a programming language - Company: Vercel (vercel.com) - Announced: 2026-08-27T07:00:00+00:00 - Category: developer-tool-release - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://vercel.com/blog/the-best-workflow-engine-is-a-programming-language - Record: https://forck.live/items/20664-the-best-workflow-engine-is-a-programming-language - Subject: v0 / AI SDK / AI Gateway Vercel introduces Workflow SDK, a framework that turns TypeScript into durable execution without requiring new infrastructure. It replaces Temporal's signals, queries, and updates with a single hook primitive and allows developers to write normal code with 'use workflow' and 'use step' directives. ### Factory AI — What it Takes for Coding Agents to Complete Large Software Tasks - Company: Factory AI (factory.com) - Announced: 2026-08-27T07:00:00+00:00 - Category: research-paper - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://factory.com/news/what-it-takes-for-coding-agents-to-complete-large-software-tasks - Record: https://forck.live/items/17876-what-it-takes-for-coding-agents-to-complete-large-software-tasks - Subject: Droid Factory Research compared single-agent and multi-role coding agent runs on 24 ProgramBench tasks. A multi-role system that first builds an executable standard of completion before implementation reproduced far more behavior than a single agent, e.g., GDAL from 36% to 90% behavioral parity, 7-Zip from 54% to 95%, and DuckDB from 34% to 80%. ### OpenAI — Expanding OpenAI’s presence in Brazil - Company: OpenAI (openai.com) - Announced: 2026-08-27T03:00:00+00:00 - Category: availability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://openai.com/index/expanding-our-presence-in-brazil - Record: https://forck.live/items/4497-expanding-openai-s-presence-in-brazil - Subject: GPT / ChatGPT / API OpenAI is expanding its presence in Brazil to increase AI adoption among developers, businesses, and communities. ### GitHub — Enterprise-managed settings now support autoUpdate for plugin marketplaces - Company: GitHub (github.com) - Announced: 2026-08-27T01:52:19+00:00 - Category: capability-change - Coverage: 1 outlet - Announcement: yes - Group: covered - Source: https://github.blog/changelog/2026-08-26-enterprise-managed-settings-now-support-autoupdate-for-plugin-marketplaces - Record: https://forck.live/items/4493-enterprise-managed-settings-now-support-autoupdate-for-plugin-marketplaces - Subject: Copilot Enterprise-managed settings now support opting individual plugin marketplaces into automatic updates by setting autoUpdate: true, reducing manual maintenance for organization customizations. This capability is generally available with Copilot Business and Copilot Enterprise in the GitHub Copilot app, Copilot CLI, and Visual Studio Code. ### Thinking Machines Lab — Putting Task Expertise into RL Achieves State-of-the-Art Performance on Text-to-SQL - Company: Thinking Machines Lab (thinkingmachines.ai) - Announced: 2026-08-27 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://thinkingmachines.ai/news/putting-task-expertise-into-rl - Record: https://forck.live/items/18553-putting-task-expertise-into-rl-achieves-state-of-the-art-performance-on-text - Subject: Thinking Machines / Inkling Thinking Machines Lab describes fine-tuning a model, ReViSQL-K2.6, that achieves human-level accuracy on text-to-SQL without scaffolding, using reinforcement learning with verifiable rewards (RLVR) on a cleaned training set called BIRD-Platinum. The model exceeds the human mark of 92.96% on the BIRD benchmark when picking from 16 samples, at a cost of $0.56 per task. The team released the cleaned training set and evaluation set to the community. ### Ai2 — Ai2 and Providence Swedish Cancer Institute partner to advance AI-assisted scientific discovery - Company: Ai2 (allenai.org) - Announced: 2026-08-27 - Category: partnership-acquisition - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://allenai.org/blog/swedish-autodiscovery-partnership - Record: https://forck.live/items/18536-ai2-and-providence-swedish-cancer-institute-partner-to-advance-ai-assisted - Subject: Olmo / Molmo / Asta Ai2 announced a partnership with the Paul G. Allen Research Center at Providence Swedish Cancer Institute to apply its AI-powered scientific discovery platform, AutoDiscovery, to cancer research datasets. The collaboration builds on a research paper where AutoDiscovery was used to analyze breast cancer data, uncovering a stronger immune signature in invasive lobular carcinoma than previously recognized, which was validated across independent datasets and laboratory analysis. The partnership also includes a local deployment of AutoDiscovery at the research center to enable analysis while keeping clinical data secure. ### Genspark — From Scattered Tools to One Platform: How This Engineer Rebuilt His Workflow with Genspark - Company: Genspark (genspark.ai) - Announced: 2026-08-27 - Category: not stated - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://www.genspark.ai/blog/from-scattered-tools-to-one-platform - Record: https://forck.live/items/18242-from-scattered-tools-to-one-platform-how-this-engineer-rebuilt-his-workflow - Subject: Genspark ### Apple — From Preferences to Principles: Rubric-Based Alignment for Grounded Knowledge Answers - Company: Apple (apple.com) - Announced: 2026-08-27 - Category: research-paper - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://machinelearning.apple.com/research/rubric-based-alignment - Record: https://forck.live/items/10305-from-preferences-to-principles-rubric-based-alignment-for-grounded-knowledge - Subject: Machine Learning Research Apple researchers introduce a rubric-based reward framework for open-domain question answering that generates query-specific rubrics grounded in retrieved evidence and decomposed into multiple quality dimensions. The approach improves over the instruction-tuned baseline by 6.5% and over flat rubric variants by 4% across three evaluation axes (composition, grounding, and instruction-following). Conditioning rubrics on retrieved evidence improves factual support, while decomposing rubrics into quality-specific dimensions further improves coherence, organization, and adherence to query requirements. ### Cartesia — Introducing Sonic-3.6 - Company: Cartesia (cartesia.ai) - Announced: 2026-08-27 - Category: model-update - Coverage: not counted - Announcement: yes - Group: models - Source: https://www.cartesia.ai/blog/sonic-3.6 - Record: https://forck.live/items/8006-introducing-sonic-3-6 - Subject: Sonic / Ink Cartesia released Sonic-3.6, an updated voice model that improves naturalness and quality, with listeners preferring it over competitors in blind head-to-head tests across multiple languages. The model supports 44 languages, 61 locales, and 500+ preset voices, including new languages Odia and Urdu. It offers sub-90ms latency and is generally available via the API and playground. ### Amp — So Long, TUI Sidebar - Company: Amp (ampcode.com) - Announced: 2026-08-27 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://ampcode.com/news/so-long-tui-sidebar - Record: https://forck.live/items/7707-so-long-tui-sidebar - Subject: Amp Amp announced it is removing the sidebar from its TUI, and recommends users manage threads and orbs in the web and native apps instead. ### Amp — Projects with Multiple Repositories - Company: Amp (ampcode.com) - Announced: 2026-08-27 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://ampcode.com/news/multi-repo-projects - Record: https://forck.live/items/7706-projects-with-multiple-repositories - Subject: Amp Amp now supports projects with multiple repositories. Additional repositories are checked out to an adjacent directory and the change panel shows diffs across all repositories. Users can add up to 20 additional repositories per project. ### LG AI Research — [ACL 2026] ReSQL: A Self-Improving Text-to-SQL Framework through Retrieval-Augmented Error Reasoning - Company: LG AI Research (lgresearch.ai) - Announced: 2026-08-27 - Category: research-paper - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://www.lgresearch.ai/blog/view?seq=676 - Record: https://forck.live/items/4736-acl-2026-resql-a-self-improving-text-to-sql-framework-through-retrieval - Subject: EXAONE LG AI Research introduces ReSQL, a self-improving Text-to-SQL framework that uses retrieval-augmented error reasoning to correct queries, demonstrating error reductions on benchmarks. ### Anthropic — Expanding our support for scientists - Company: Anthropic (anthropic.com) - Announced: 2026-08-27 - Category: availability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://www.anthropic.com/news/expanding-support-for-scientists - Record: https://forck.live/items/4527-expanding-our-support-for-scientists - Subject: Claude Anthropic announces 10,000 free or discounted Claude subscriptions for scientists via a new team plan, expands AI for Science program with up to $50,000 in credits per project, and notes restrictions on Opus, Fable, and Mythos models for certain fields. ### Anthropic — Previewing the Model Hardware Standard - Company: Anthropic (anthropic.com) - Announced: 2026-08-27 - Category: infrastructure-release - Coverage: 18 outlets - Announcement: yes - Group: covered - Source: https://www.anthropic.com/news/model-hardware-standard-research-preview - Record: https://forck.live/items/4524-previewing-the-model-hardware-standard - Subject: Claude Anthropic is previewing the Model Hardware Standard (MHS), a shared specification enabling AI agents to safely operate physical devices like microscopes and robotic arms, reducing integration time from weeks to hours or minutes. ### Cursor — Start from scratch, without a repo - Company: Cursor (cursor.com) - Announced: 2026-08-27 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://cursor.com/changelog/start-from-scratch - Record: https://forck.live/items/4521-start-from-scratch-without-a-repo - Subject: Cursor Cursor Cloud Agents now support starting without a connected repository, allowing users to prompt from scratch and save work to a Cursor Origin repo. ### Cohere — Why forward-deployed engineers should build capability, not dependency - Company: Cohere (cohere.com) - Announced: 2026-08-27 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://cohere.com/blog/forward-deployed-engineers-capability-building - Record: https://forck.live/items/4507-why-forward-deployed-engineers-should-build-capability-not-dependency - Subject: Command Cohere discusses how forward-deployed engineers (FDEs) can build customer capability to avoid vendor lock-in, emphasizing co-building and knowledge transfer rather than creating operational dependency. ### Cohere — Introducing Parse: Enterprise document intelligence at scale - Company: Cohere (cohere.com) - Announced: 2026-08-27 - Category: availability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://cohere.com/blog/parse - Record: https://forck.live/items/4496-introducing-parse-enterprise-document-intelligence-at-scale - Subject: Command Cohere announces Parse, a cost-effective vision language model for enterprise document parsing, available via API and Model Vault, with pricing at $1.50 per 1,000 pages. ### Replit — Intelligent Model Routing on Replit - Company: Replit (replit.com) - Announced: 2026-08-26T22:18:53+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://replit.com/blog/intelligent-model-routing - Record: https://forck.live/items/17884-intelligent-model-routing-on-replit - Subject: Replit Agent Replit introduced Intelligent Model Routing, which automatically selects the best model for each task based on quality, speed, and cost. The feature is available to all users, with Free Mode as the default and notifications when higher-powered modes may incur costs. Core and Pro users can still manually select models, and enterprise administrators can define an approved model set for their workspace. ### GitHub — Global model policy generally available - Company: GitHub (github.com) - Announced: 2026-08-26T22:08:38+00:00 - Category: availability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://github.blog/changelog/2026-08-26-global-model-policy-generally-available - Record: https://forck.live/items/4486-global-model-policy-generally-available - Subject: Copilot GitHub is rolling out enforcement of a default model policy for Copilot Business and Enterprise plans, making it generally available. Models not explicitly configured will follow the policy, and open-weight models are disabled by default. ### NVIDIA — NVIDIA NVLink Fusion Expands With NVHBM Custom High-Bandwidth Memory - Company: NVIDIA (nvidia.com) - Announced: 2026-08-26T21:05:30+00:00 - Category: infrastructure-release - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://blogs.nvidia.com/blog/nvlink-fusion-nvhbm-custom-high-bandwidth-memory/ - Record: https://forck.live/items/4479-nvidia-nvlink-fusion-expands-with-nvhbm-custom-high-bandwidth-memory - Subject: AI platform NVIDIA announced NVHBM, a custom high-bandwidth memory technology that integrates the memory controller into the HBM stack, delivering up to 30% greater memory bandwidth, 15% lower power consumption, and freeing up to 25% more area on XPU compute die. It expands the NVLink Fusion platform and will be validated by memory partners. Amazon's Annapurna Labs will be the first to work on NVHBM. ### Runway — The Kinetix Team is Joining Runway - Company: Runway (runway.com) - Announced: 2026-08-26T19:37:32.047000+00:00 - Category: partnership-acquisition - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://runway.com/news/company-news/kinetix-joins-runway - Record: https://forck.live/items/9261-the-kinetix-team-is-joining-runway - Subject: Gen / Aleph Runway announced that the team from Kinetix, a Paris-based AI research lab, is joining the company. The Kinetix team's work on 3D human motion and physically-grounded video generation will support Runway's research into world models for robotics, and they will join Runway's Paris hub. ### Amazon — Evaluate any agent framework with Amazon Bedrock AgentCore Evaluations - Company: Amazon (amazon.com) - Announced: 2026-08-26T19:13:35+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://aws.amazon.com/blogs/machine-learning/evaluate-any-agent-framework-with-amazon-bedrock-agentcore-evaluations/ - Record: https://forck.live/items/4475-evaluate-any-agent-framework-with-amazon-bedrock-agentcore-evaluations - Subject: Bedrock / Nova Amazon Bedrock AgentCore Evaluations is a new evaluation service that uses OpenTelemetry to score any agent framework (LangGraph, LlamaIndex, OpenAI Agents SDK, Google ADK, Claude Agent SDK, Strands Agents) running on Amazon Bedrock AgentCore runtime, without requiring framework-specific integration. It reads three span roles (invoke agent, inference, execute tool) to reconstruct sessions and apply evaluators such as GoalSuccessRate, Correctness, Helpfulness, and custom LLM-as-a-judge. The service bridges both OpenTelemetry GenAI conventions and OpenInference specifications. ### Google Research — GlucoFM: Foundation model for continuous glucose monitoring - Company: Google Research (research.google) - Announced: 2026-08-26T18:42:43+00:00 - Category: new-model - Coverage: not counted - Announcement: yes - Group: models - Source: https://research.google/blog/glucofm-foundation-model-for-continuous-glucose-monitoring/ - Record: https://forck.live/items/4471-glucofm-foundation-model-for-continuous-glucose-monitoring - Subject: Research Google Research announces GlucoFM, a lightweight self-supervised foundation model for continuous glucose monitoring that uses a dual-stream design to separate slower glycemic trends from short-term deviations, achieving state-of-the-art performance on metabolic prediction tasks including diabetes risk assessment, insulin resistance, and beta-cell dysfunction across multiple cohorts and devices. ### Google DeepMind — Intelligent transcription with Gemini 3.5 Transcribe - Company: Google DeepMind (deepmind.google) - Announced: 2026-08-26T17:01:00+00:00 - Category: new-model - Coverage: not counted - Announcement: yes - Group: models - Source: https://deepmind.google/blog/intelligent-transcription-with-gemini-3-5-transcribe/ - Record: https://forck.live/items/4467-intelligent-transcription-with-gemini-3-5-transcribe - Subject: Gemini Google DeepMind announces Gemini 3.5 Transcribe, a new speech-to-text model for real-time transcription, available via the Gemini API in Google AI Studio and Gemini Enterprise Agent Platform. The model offers smart transcription, function calling, custom vocabulary, multilingual support, and multi-speaker identification. It is accessible through streaming (Live API with gemini-3.5-transcribe-live) and pre-recorded audio (Interactions API with gemini-3.5-transcribe) APIs. ### Amazon — How GoDaddy transformed its analytics with Amazon Quick - Company: Amazon (amazon.com) - Announced: 2026-08-26T16:51:53+00:00 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://aws.amazon.com/blogs/machine-learning/how-godaddy-transformed-its-analytics-with-amazon-quick/ - Record: https://forck.live/items/4464-how-godaddy-transformed-its-analytics-with-amazon-quick - Subject: Bedrock / Nova GoDaddy migrated from a legacy BI tool to Amazon Quick Sight, achieving performance improvements and cost savings. It is a customer success story, not a product or model announcement. ### Amazon — Natera’s intelligent appointment scheduling with Amazon Bedrock AgentCore - Company: Amazon (amazon.com) - Announced: 2026-08-26T16:36:35+00:00 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://aws.amazon.com/blogs/machine-learning/nateras-intelligent-appointment-scheduling-with-amazon-bedrock-agentcore/ - Record: https://forck.live/items/4465-natera-s-intelligent-appointment-scheduling-with-amazon-bedrock-agentcore - Subject: Bedrock / Nova Natera built an intelligent appointment scheduling voice agent using Amazon Bedrock AgentCore, achieving 100% tool-calling accuracy and sub-7-second latency at less than $0.01 per call. ### Amazon — Bring your own model with Amazon SageMaker AI: Script mode in SDK v3 - Company: Amazon (amazon.com) - Announced: 2026-08-26T16:31:32+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://aws.amazon.com/blogs/machine-learning/bring-your-own-model-with-amazon-sagemaker-ai-script-mode-in-sdk-v3/ - Record: https://forck.live/items/4466-bring-your-own-model-with-amazon-sagemaker-ai-script-mode-in-sdk-v3 - Subject: Bedrock / Nova Amazon announces script mode in SageMaker Python SDK v3, which allows users to bring their own model without building Docker images, using new unified classes ModelTrainer and ModelBuilder, and a SourceCode object for runtime code injection. ### Amazon — Preparing data for supervised fine-tuning Part 2: Advanced data strategies - Company: Amazon (amazon.com) - Announced: 2026-08-26T16:24:05+00:00 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://aws.amazon.com/blogs/machine-learning/preparing-data-for-supervised-fine-tuning-part-2-advanced-data-strategies/ - Record: https://forck.live/items/4443-preparing-data-for-supervised-fine-tuning-part-2-advanced-data-strategies - Subject: Bedrock / Nova Advanced data preparation strategies for supervised fine-tuning, covering learning curve analysis, data subset selection, augmentation, and mixing, with references to Amazon Nova customization findings. ### Amazon — Preparing data for supervised fine-tuning Part 1: Formatting and quality - Company: Amazon (amazon.com) - Announced: 2026-08-26T16:24:02+00:00 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://aws.amazon.com/blogs/machine-learning/preparing-data-for-supervised-fine-tuning-part-1-formatting-and-quality/ - Record: https://forck.live/items/4444-preparing-data-for-supervised-fine-tuning-part-1-formatting-and-quality - Subject: Bedrock / Nova Guidance on preparing data for supervised fine-tuning (SFT) of foundation models. It covers data quality checks, formatting, and train/evaluation splits, with references to Amazon Nova and Amazon Bedrock as examples, but does not announce any new product, model, or feature. ### Microsoft — The Economics of Agent Optimization: Four ways to lower the cost - Company: Microsoft (microsoft.com) - Announced: 2026-08-26T16:00:00+00:00 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://azure.microsoft.com/en-us/blog/the-economics-of-agent-optimization-four-ways-to-lower-the-cost/ - Record: https://forck.live/items/4492-the-economics-of-agent-optimization-four-ways-to-lower-the-cost - Subject: Azure AI Four strategies to optimize agent costs: selecting the right model per request, caching, prompt and agent optimization, and observability. It describes capabilities like model router, deployment options, fine-tuning, and prompt optimizer to reduce cost per successful outcome. ### Amazon — Connect Amazon Bedrock AgentCore to cross-account knowledge bases - Company: Amazon (amazon.com) - Announced: 2026-08-26T15:48:10+00:00 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://aws.amazon.com/blogs/machine-learning/connect-amazon-bedrock-agentcore-to-cross-account-knowledge-bases/ - Record: https://forck.live/items/4441-connect-amazon-bedrock-agentcore-to-cross-account-knowledge-bases - Subject: Bedrock / Nova To connect Amazon Bedrock AgentCore agents in one AWS account to knowledge bases in another account using Amazon Redshift Serverless, with two orchestration variants: a code-based Strands agent and a declarative AgentCore harness, without copying source data across accounts. The post covers architecture, flow, and selection criteria for the two generally available AgentCore orchestration models. Models used include Amazon Nova Pro and Claude Haiku 4.5. The solution relies on cross-account IAM roles and AWS STS to assume role in the knowledge base account and call RetrieveAndGenerate. The linked GitHub sample provides deployment procedures for both variants. The post also advises on when to use a simpler approach versus an agent for cross-account queries. ### Google DeepMind — The mathematics of AI uncertainty - Company: Google DeepMind (deepmind.google) - Announced: 2026-08-26T15:45:28+00:00 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://www.youtube.com/watch?v=tBjgCj_dGZM - Record: https://forck.live/items/19941-the-mathematics-of-ai-uncertainty Google DeepMind published a video interview with Zoubin Ghahramani, a VP of research at the lab and Cambridge professor, discussing his three decades of work on uncertainty in AI and its potential role in future AI progress. ### Harvey — Memory is Here: Harvey, Personalized - Company: Harvey (harvey.ai) - Announced: 2026-08-26T14:20:00+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://www.harvey.ai/blog/memory-is-here-harvey-personalized - Record: https://forck.live/items/17530-memory-is-here-harvey-personalized - Subject: Harvey Harvey introduced Memory, a feature that lets users set persistent preferences—such as drafting style, jurisdictional distinctions, or word count limits—that apply automatically across the Web App, Outlook, and Word Add-In. Memories are saved only when explicitly requested or when a prompt signals a preference, and users can review, edit, search, or delete them in Personal Settings. The feature is available in Early Access, with general availability to follow. ### Google Cloud — FinOps for the AI era: New flexible billing and cost controls for agents - Company: Google Cloud (cloud.google.com) - Announced: 2026-08-26T13:30:00+00:00 - Category: pricing-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://cloud.google.com/blog/products/ai-machine-learning/flexible-billing-and-cost-controls-for-agents-on-google-cloud/ - Record: https://forck.live/items/17335-finops-for-the-ai-era-new-flexible-billing-and-cost-controls-for-agents - Subject: Gemini Enterprise Google Cloud introduced expanded billing flexibility and cost management tools for agent workloads in Gemini Enterprise. New options include a pay-as-you-go consumption edition, Flexible Savings Plans offering 10–20% discounts on token costs, consolidated pooled quotas, and deferred execution pricing that can cut inference costs by up to half. Google Antigravity and Android Studio AI usage are now included in Gemini Enterprise subscriptions for select customers, rolling out broadly soon. ### OpenAI — Learning never stops: How AI makes learning continuous - Company: OpenAI (openai.com) - Announced: 2026-08-26T10:00:00+00:00 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://openai.com/index/learning-never-stops - Record: https://forck.live/items/4470-learning-never-stops-how-ai-makes-learning-continuous - Subject: GPT / ChatGPT / API OpenAI published a report about how students and educators use ChatGPT to make learning more continuous, extending beyond the classroom. ### OpenAI — Bringing ChatGPT for Teachers to more U.S. school districts - Company: OpenAI (openai.com) - Announced: 2026-08-26T10:00:00+00:00 - Category: availability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://openai.com/index/bringing-chatgpt-for-teachers-to-more-us-school-districts - Record: https://forck.live/items/4469-bringing-chatgpt-for-teachers-to-more-u-s-school-districts - Subject: GPT / ChatGPT / API OpenAI is expanding ChatGPT for Teachers to 55 U.S. school systems, providing secure AI tools, training, and support to over 100,000 educators and staff. ### Google Antigravity — Visualizing with the help of Antigravity - Company: Google Antigravity (antigravity.google) - Announced: 2026-08-26T08:00:00+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://antigravity.google/blog/visualizing-with-the-help-of-antigravity - Record: https://forck.live/items/4540-visualizing-with-the-help-of-antigravity - Subject: Antigravity Google Antigravity announces Generative UI and visual tooling that allows agents to create, render, and iterate on rich visual components, dynamic dashboards, interactive artifacts, and design mockups in real time. ### Perceptron — Introducing Isaac 0.5 - Company: Perceptron (perceptron.inc) - Announced: 2026-08-26 - Category: new-model - Coverage: not counted - Announcement: yes - Group: models - Source: https://www.perceptron.inc/blog/introducing-isaac-0-5 - Record: https://forck.live/items/18608-introducing-isaac-0-5 - Subject: Perceptron Isaac / Perceptron Mk1 Perceptron released Isaac 0.5, an open-source embodied foundation model with 36 billion parameters. The model unifies multimodal video understanding, embodied reasoning, and robot control, and establishes a scaling law showing that increasing general video data can dramatically reduce the amount of teleoperation data needed to achieve a given action loss. Isaac 0.5 is released as a complete system including checkpoints, training code, and inference code via LeRobot. ### Ai2 — How researchers adapted Dolma for better Thai language models - Company: Ai2 (allenai.org) - Announced: 2026-08-26 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://allenai.org/blog/thai-llm-dolma - Record: https://forck.live/items/18537-how-researchers-adapted-dolma-for-better-thai-language-models - Subject: Olmo / Molmo / Asta Ai2 reported that Thai researchers used its open Dolma toolkit to create Mangosteen, a 47-billion-token Thai pretraining corpus. The team modified Dolma's pipeline to handle Thai's sentence boundaries and quality filters, and found that their curation removed over 80% of Common Crawl data and nearly half of FineWeb2 while maintaining or improving model performance. ### Apple — Luce: Relightable Gaussians for 3D Asset Generation - Company: Apple (apple.com) - Announced: 2026-08-26 - Category: research-paper - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://machinelearning.apple.com/research/relightable-gaussians-3d-generation - Record: https://forck.live/items/10308-luce-relightable-gaussians-for-3d-asset-generation - Subject: Machine Learning Research Apple researchers propose Luce, a 3D representation that unifies geometry and physically based rendering (PBR) materials within a voxelized multimodal Gaussian cloud. A rectified-flow transformer generates this representation from a single image, and the latent decodes into relightable PBR Gaussians and an optional textured mesh. On the Toys4K benchmark, Luce achieves state-of-the-art single-image-to-3D generation, improving FID by 28% over the strongest baseline. ### Apple — IDEA Prune: An Integrated Enlarge-and-Prune Pipeline in Generative Language Model Pretraining - Company: Apple (apple.com) - Announced: 2026-08-26 - Category: research-paper - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://machinelearning.apple.com/research/idea-prune-pipeline - Record: https://forck.live/items/10307-idea-prune-an-integrated-enlarge-and-prune-pipeline-in-generative-language - Subject: Machine Learning Research Apple researchers propose an integrated enlarge-and-prune pipeline for generative language model pretraining that combines enlarged model training, pruning, and recovery under a single cosine annealing learning rate schedule, complemented by a novel iterative structured pruning method. Experiments compressing 2.8B models to 1.3B with up to 2T tokens show the approach mitigates knowledge loss and achieves superior pruned model performance. ### Apple — PROOF-Gen: From Optimized Data to Better Distillation - Company: Apple (apple.com) - Announced: 2026-08-26 - Category: research-paper - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://machinelearning.apple.com/research/proof-gen-optimized-distillation - Record: https://forck.live/items/10306-proof-gen-from-optimized-data-to-better-distillation - Subject: Machine Learning Research Apple researchers introduced PROOF-Gen, a method that recovers golden trajectories from failed teacher generations via per-scenario prompt optimization, improving tool-calling distillation. On τ2-bench it recovered 93% of failed scenarios; fine-tuning Qwen3-4B-Instruct-2507 improved Pass@1 from 0.132 to 0.529, and Gemma 4 E4B-it gained +7.2pp on BFCL v4 multi-turn. In a deployed pipeline, the method lifted goal completion by +6.3pp and transferred to an on-device model with positive results across locales. ### Anthropic — How Warp builds self-improving agents on Claude - Company: Anthropic (anthropic.com) - Announced: 2026-08-26 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://claude.com/blog/how-warp-builds-self-improving-agents-on-claude - Record: https://forck.live/items/7415-how-warp-builds-self-improving-agents-on-claude - Subject: Claude Warp developed a self-improving agent framework using skills on the Claude Platform, consisting of a base skill for domain knowledge and an improver skill that uses human feedback to refine the base skill via file-based updates. ### Anthropic — Claude gets its own browser in Cowork - Company: Anthropic (anthropic.com) - Announced: 2026-08-26 - Category: capability-change - Coverage: 5 outlets - Announcement: yes - Group: covered - Source: https://claude.com/blog/cowork-built-in-browser - Record: https://forck.live/items/7402-claude-gets-its-own-browser-in-cowork - Subject: Claude Anthropic announced that Claude now has a built-in browser in the Claude Cowork desktop app, enabling Claude to navigate websites, read pages, and fill forms without requiring the user's own browser. The feature is rolling out to Pro, Max, and Team plans on desktop, with Enterprise admin control available. The built-in browser is separate from the user's browser and does not access user tabs, bookmarks, or passwords. ### Anthropic — Claude in Chrome is generally available - Company: Anthropic (anthropic.com) - Announced: 2026-08-26 - Category: availability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://claude.com/blog/claude-in-chrome-generally-available - Record: https://forck.live/items/7401-claude-in-chrome-is-generally-available - Subject: Claude Anthropic made Claude in Chrome generally available on every paid Claude plan. Claude can now take autonomous actions in the browser with a safety classifier that validates each action before it runs. The company reported evaluation results showing improved defenses against prompt injection attacks across multiple models. ### xAI — Grok 4.6 on Microsoft Foundry - Company: xAI (x.ai) - Announced: 2026-08-26 - Category: availability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://x.ai/news/grok-4-6-microsoft-foundry - Record: https://forck.live/items/4491-grok-4-6-on-microsoft-foundry - Subject: Grok xAI announces that Grok 4.6 is now available on Microsoft Foundry, offering a 500k context window and configurable reasoning efforts. ### OpenAI — The Hugging Face incident and the road ahead - Company: OpenAI (openai.com) - Announced: 2026-08-26 - Category: not stated - Coverage: 14 outlets - Announcement: no - Group: covered - Source: https://openai.com/index/hugging-face-incident-and-the-road-ahead - Record: https://forck.live/items/4477-the-hugging-face-incident-and-the-road-ahead - Subject: GPT / ChatGPT / API OpenAI shares findings from a security incident involving Hugging Face and outlines steps to strengthen AI model security, monitoring, and alignment. ### Hugging Face — Training and Finetuning Multi-Vector Embedding Models with Sentence Transformers - Company: Hugging Face (huggingface.co) - Announced: 2026-08-26 - Category: developer-tool-release - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://huggingface.co/blog/train-multi-vector-encoder - Record: https://forck.live/items/4436-training-and-finetuning-multi-vector-embedding-models-with-sentence-transformers - Subject: Platform Sentence Transformers v6.0 introduces a new model type, MultiVectorEncoder, for ColBERT-style late interaction retrieval, along with a complete training approach. The blog post demonstrates how to finetune multi-vector models that outperform general-purpose retrievers on domain-specific data, and shows that a finetuned model (multi-vector-encoder/mLateOn-medical) trained in 14.5 hours on a single RTX 3090 outperforms all general-purpose retrieval models on a medical retrieval evaluation. ### OpenAI — How loveholidays is making everyone a builder with Codex - Company: OpenAI (openai.com) - Announced: 2026-08-26 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://openai.com/index/loveholidays - Record: https://forck.live/items/4428-how-loveholidays-is-making-everyone-a-builder-with-codex - Subject: GPT / ChatGPT / API Loveholidays uses OpenAI Codex to enable software development across the business, allowing teams to quickly turn ideas into products. ### GitHub — GitHub Copilot app Customize tab is generally available - Company: GitHub (github.com) - Announced: 2026-08-25T20:05:26+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://github.blog/changelog/2026-08-25-github-copilot-app-customize-tab-is-generally-available - Record: https://forck.live/items/4419-github-copilot-app-customize-tab-is-generally-available - Subject: Copilot GitHub announced the general availability of the Customize tab in the GitHub Copilot app, which unifies MCP servers, plugins, skills, and canvases in a single interface for discovering and managing customizations. ### Google Research — AgentHands: Generating interactive hand gestures for spatially grounded agent conversations in XR - Company: Google Research (research.google) - Announced: 2026-08-25T19:10:59+00:00 - Category: research-paper - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://research.google/blog/agenthands-generating-interactive-hand-gestures-for-spatially-grounded-agent-conversations-in-xr/ - Record: https://forck.live/items/4418-agenthands-generating-interactive-hand-gestures-for-spatially-grounded-agent - Subject: Research AgentHands is a research prototype that uses LLMs to generate synchronized hand gestures for AI agents in XR, enabling spatially grounded conversations. ### Amazon — Agentic observability with Amazon OpenSearch Service MCP Apps - Company: Amazon (amazon.com) - Announced: 2026-08-25T19:00:09+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://aws.amazon.com/blogs/machine-learning/agentic-observability-with-amazon-opensearch-service-mcp-apps/ - Record: https://forck.live/items/4417-agentic-observability-with-amazon-opensearch-service-mcp-apps - Subject: Bedrock / Nova Amazon OpenSearch Service announced MCP Apps, a capability that extends the Model Context Protocol to return interactive visualizations alongside text responses in AI assistant chats, enabling verification without switching tabs. ### Amazon — Governed reports with Amazon Quick Desktop and Amazon FSx for NetApp ONTAP - Company: Amazon (amazon.com) - Announced: 2026-08-25T16:35:01+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://aws.amazon.com/blogs/machine-learning/governed-reports-with-amazon-quick-desktop-and-amazon-fsx-for-netapp-ontap/ - Record: https://forck.live/items/4413-governed-reports-with-amazon-quick-desktop-and-amazon-fsx-for-netapp-ontap - Subject: Bedrock / Nova Amazon announces Amazon Quick Desktop, a native desktop application that extends Amazon Quick with AI-assisted reporting capabilities, including local file access, background processing, and a personal knowledge graph. The post describes a governed weekly reporting workflow using Amazon Quick Desktop, Amazon Quick skills, Amazon FSx for NetApp ONTAP, Amazon S3 access points, and Slack. ### Hugging Face — Granite 4.2 LLMs: How They're Built - Company: Hugging Face (huggingface.co) - Announced: 2026-08-25T15:14:14+00:00 - Category: new-model - Coverage: not counted - Announcement: no - Group: routine - Source: https://huggingface.co/blog/ibm-granite/granite-4-2 - Record: https://forck.live/items/4407-granite-4-2-llms-how-they-re-built - Subject: Platform IBM announced the Granite 4.2 family of reasoning LLMs in three sizes (3B, 8B, 30B), pre-trained from scratch on about 15T tokens, with a context window extended to 512K tokens, supervised fine-tuning on reasoning and agentic data, and released under the Apache 2.0 license. ### Hugging Face — Extremely Fast and Accurate Transcription with Granite Speech 5.0 Turbo CTC - Company: Hugging Face (huggingface.co) - Announced: 2026-08-25T15:01:00+00:00 - Category: new-model - Coverage: not counted - Announcement: yes - Group: models - Source: https://huggingface.co/blog/ibm-granite/granite-speech-5-0-470m-turboctc - Record: https://forck.live/items/4405-extremely-fast-and-accurate-transcription-with-granite-speech-5-0-turbo-ctc - Subject: Platform IBM announces the release of two new Granite Speech 5.0 Turbo CTC speech recognition models: a 470M-parameter Apache 2.0 licensed model and a CC-BY-NC-SA-4.0 licensed model, both offering high accuracy and over 12,600 RTFx throughput. ### Stability AI — The Entertainment Industry’s Biggest Names Back Stability AI in Latest Funding Round - Company: Stability AI (stability.ai) - Announced: 2026-08-25T14:59:22+00:00 - Category: funding-company-news - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://stability.ai/news-updates/stability-ai-latest-funding-backed-by-entertainment-industry-biggest-names - Record: https://forck.live/items/17473-the-entertainment-industry-s-biggest-names-back-stability-ai-in-latest-funding - Subject: Stable Diffusion Stability AI announced a $76M Series B fundraise, bringing total funding to $232M. The round includes new investors Electronic Arts, Sony Music Group, Universal Music Group, and Warner Music Group, alongside existing backers. The capital will support product development, applied research, and professional services expansion. ### Google Cloud — Now introducing Gemini Enterprise for Legal - Company: Google Cloud (cloud.google.com) - Announced: 2026-08-25T12:00:00+00:00 - Category: product-launch - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://cloud.google.com/blog/products/ai-machine-learning/introducing-gemini-enterprise-for-legal/ - Record: https://forck.live/items/17336-now-introducing-gemini-enterprise-for-legal - Subject: Gemini Enterprise Google Cloud introduced Gemini Enterprise for Legal, a governed AI platform for law firms and corporate legal departments. The platform combines purpose-built skills for legal tasks, secure connectors to existing systems, pre-built and customizable agents, and an open partner ecosystem, all managed through a centralized control plane that enforces security policies and data isolation. ### Hugging Face — Quantization-Aware Healing: a compressed, 4-bit model that outperforms its full-precision original - Company: Hugging Face (huggingface.co) - Announced: 2026-08-25T11:39:24+00:00 - Category: research-paper - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://huggingface.co/blog/MultiverseComputingCAI/quantization-aware-healing - Record: https://forck.live/items/4396-quantization-aware-healing-a-compressed-4-bit-model-that-outperforms-its-full - Subject: Platform Introduces Quantization-Aware Healing (QAH), a method that distills from the original pre-compression model to recover capabilities lost during structural compression and quantization. Applied to a GPT-OSS 120B model compressed to 60B and quantized to MXFP4, it outperforms its own full-precision version on 7 of 9 benchmarks. ### Lovable — How we made Lovable apps work with the rest of your stack - Company: Lovable (lovable.dev) - Announced: 2026-08-25T09:00:00+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://lovable.dev/blog/how-we-made-lovable-apps-work-with-the-rest-of-your-stack - Record: https://forck.live/items/12057-how-we-made-lovable-apps-work-with-the-rest-of-your-stack - Subject: Lovable / AI app builder Lovable built a system of app connectors that allow users to integrate their AI-generated apps with external services and data sources. The company developed a Connector Gateway architecture to manage OAuth credentials securely, handle token refresh logic, and enable integrations with nearly 100 services including Gmail, Slack, Firecrawl, ElevenLabs, and Perplexity. ### OpenAI — The full stack behind abundant intelligence - Company: OpenAI (openai.com) - Announced: 2026-08-25T07:05:00+00:00 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://openai.com/index/the-full-stack-behind-abundant-intelligence - Record: https://forck.live/items/4410-the-full-stack-behind-abundant-intelligence - Subject: GPT / ChatGPT / API OpenAI's CFO discusses the compounding effects of advances across hardware, compute, models, and products to deliver more useful intelligence at greater scale and lower cost. ### OpenAI — Jalapeño’s first results show industry-leading speed and efficiency in AI inference - Company: OpenAI (openai.com) - Announced: 2026-08-25T07:00:00+00:00 - Category: infrastructure-release - Coverage: 13 outlets - Announcement: yes - Group: covered - Source: https://openai.com/index/jalapeno-first-results - Record: https://forck.live/items/4404-jalape-o-s-first-results-show-industry-leading-speed-and-efficiency-in-ai - Subject: GPT / ChatGPT / API OpenAI announced Jalapeño, a custom inference chip that offers faster, more power-efficient inference with higher throughput and lower latency for modern models. ### Vercel — The end of credential sprawl for agents - Company: Vercel (vercel.com) - Announced: 2026-08-25T04:00:00+00:00 - Category: availability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://vercel.com/blog/the-end-of-credential-sprawl-for-agents - Record: https://forck.live/items/20666-the-end-of-credential-sprawl-for-agents - Subject: v0 / AI SDK / AI Gateway Vercel announced the general availability of Vercel Connect, a credential management system that replaces long-lived tokens with short-lived tokens requested at runtime, scoped per task, and automatically refreshed. The platform now includes over 100 preset connectors for services like Slack, GitHub, and Snowflake, and adds governance features such as fine-grained RBAC, audit logs, and token observability. Vercel Connect is available across custom environments, in the eve agent framework, and in v0 builds. ### Vercel — Introducing Run SDK: secure eval for your agents - Company: Vercel (vercel.com) - Announced: 2026-08-25T04:00:00+00:00 - Category: developer-tool-release - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://vercel.com/blog/introducing-run - Record: https://forck.live/items/20665-introducing-run-sdk-secure-eval-for-your-agents - Subject: v0 / AI SDK / AI Gateway Vercel released the Run SDK, a package for executing untrusted JavaScript and TypeScript in a sandboxed QuickJS context without direct access to the host application or system. Applications expose narrow host functions and can interrupt execution for human-in-the-loop approval or authentication, then resume without repeating completed work. The SDK powers code mode tool execution in the AI SDK and is intended for JavaScript computation inside an application. ### SenseTime — 我是全新的商汤小浣熊，今天的工作先交给我！ - Company: SenseTime (sensetime.com) - Announced: 2026-08-25 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://www.sensetime.com/cn/news/-2 - Record: https://forck.live/items/18492- - Title in English (translated by forck.live from zh): I am the new SenseTime Raccoon, leave today's work to me! - Subject: SenseNova / 商汤 SenseTime rebranded its 'Little Raccoon' family of AI assistants as 'SenseTime Raccoon Work', unifying the office and code variants under one brand. It also announced upcoming industry-specific editions for education, finance, and government, with further details to be released later. ### Kyutai — Pocket TTS training code released - Company: Kyutai (kyutai.org) - Announced: 2026-08-25 - Category: developer-tool-release - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://kyutai.org/blog/2026-08-25-pocket-tts-training - Record: https://forck.live/items/18336-pocket-tts-training-code-released - Subject: Moshi / Pocket TTS Kyutai released the training code for Pocket TTS, enabling anyone with a GPU and speech data to train their own model. ### Factory AI — Connect Droid to the tools that run your company - Company: Factory AI (factory.com) - Announced: 2026-08-25 - Category: capability-change - Coverage: not counted - Announcement: no - Group: routine - Source: https://factory.com/news/personas-and-connectors - Record: https://forck.live/items/17867-connect-droid-to-the-tools-that-run-your-company - Subject: Droid Factory AI launched Connectors, giving its Droid agent managed access to apps like Linear, Slack, Sentry, Notion, and Salesforce, and introduced Personas that pre-configure the relevant Connectors for each role (Engineering, Product, Design, etc.). Enterprise Owners and Managers can control which Connectors and Personas are available and recommend approved Connectors per role. ### Perplexity — A Local-First Agent for Private Knowledge Work - Company: Perplexity (perplexity.ai) - Announced: 2026-08-25 - Category: not stated - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://www.perplexity.ai/hub/blog/a-local-first-agent-for-private-and-cost-effective-knowledge-work - Record: https://forck.live/items/13441-a-local-first-agent-for-private-knowledge-work - Subject: Perplexity Perplexity Portable Computer is a local-first agent. The entire stack runs locally by default. ### Perplexity — Introducing Portable Computer - Company: Perplexity (perplexity.ai) - Announced: 2026-08-25 - Category: product-launch - Coverage: 13 outlets - Announcement: yes - Group: covered - Source: https://www.perplexity.ai/hub/blog/introducing-portable-computer-for-local-first-ai - Record: https://forck.live/items/10740-introducing-portable-computer - Subject: Perplexity Perplexity launched Portable Computer, a version of its AI agent that runs entirely on a local machine, built with NVIDIA for the NVIDIA DGX Spark. It uses local models (Qwen 3.8 27B or PPLX 27B) and can escalate tasks to the cloud with user permission. Private data stays on-device, and local work does not consume credits. ### Amp — Setup Without a Commit - Company: Amp (ampcode.com) - Announced: 2026-08-25 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://ampcode.com/news/setup-without-a-commit - Record: https://forck.live/items/7708-setup-without-a-commit - Subject: Amp Amp introduced the ability to store pre-clone and pre-setup scripts in project settings, allowing orbs to be set up without committing setup files to the repository. ### Anthropic — Claude's memory works everywhere, and you decide what's in it - Company: Anthropic (anthropic.com) - Announced: 2026-08-25 - Category: capability-change - Coverage: 14 outlets - Announcement: yes - Group: covered - Source: https://claude.com/blog/claudes-memory-works-everywhere-and-you-decide-whats-in-it - Record: https://forck.live/items/7417-claude-s-memory-works-everywhere-and-you-decide-what-s-in-it - Subject: Claude Anthropic announced that Claude's memory now works across both chat and Claude Cowork, allowing users to see, edit, and delete memories. Users can also enable memory for sensitive topics like health or beliefs, while default memory excludes sensitive information. ### Anthropic — Bain & Company joins the Claude Partner Network as a Global Premier partner - Company: Anthropic (anthropic.com) - Announced: 2026-08-25 - Category: partnership-acquisition - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://claude.com/blog/bain-company-joins-the-claude-partner-network-as-a-global-premier-partner - Record: https://forck.live/items/7416-bain-company-joins-the-claude-partner-network-as-a-global-premier-partner - Subject: Claude Anthropic and Bain & Company announced a global partnership, with Bain joining the Claude Partner Network as a Global Premier partner. The two firms will collaborate on client engagements in AI strategy, technology modernization, and AI-enabled operations, building on Bain's deployment of Claude to its 19,000 employees. ### OpenAI — Introducing the Admin plugin for ChatGPT Work and Codex - Company: OpenAI (openai.com) - Announced: 2026-08-25 - Category: capability-change - Coverage: 1 outlet - Announcement: yes - Group: covered - Source: https://openai.com/index/introducing-admin-plugin - Record: https://forck.live/items/4414-introducing-the-admin-plugin-for-chatgpt-work-and-codex - Subject: GPT / ChatGPT / API OpenAI introduces the Admin plugin for ChatGPT Work and Codex, enabling workspace usage analysis, member and permission management, limit adjustments, and handling of admin requests. ### Anthropic — Funding better evaluations of AI’s impact on wellbeing - Company: Anthropic (anthropic.com) - Announced: 2026-08-25 - Category: funding-company-news - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://www.anthropic.com/news/wellbeing-research-grants - Record: https://forck.live/items/4412-funding-better-evaluations-of-ai-s-impact-on-wellbeing - Subject: Claude Anthropic announces a $5 million grant program to fund independent research into how AI impacts users' wellbeing, providing direct funding, model access, and technical support to grantees building open-source evaluations. ### Cohere — The state of sovereign AI adoption in 2026 - Company: Cohere (cohere.com) - Announced: 2026-08-25 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://cohere.com/blog/state-of-sovereign-ai-adoption-2026 - Record: https://forck.live/items/4397-the-state-of-sovereign-ai-adoption-in-2026 - Subject: Command Cohere announces findings from a commissioned IDC study on sovereign AI adoption, highlighting definitional gaps, top concerns (data leakage and compliance), and the need for a holistic strategy. The post presents Cohere's private deployment architecture and North platform as solutions but does not introduce a new product, model, or feature. ### OpenAI — Disrupting a new covert influence campaign from Russia - Company: OpenAI (openai.com) - Announced: 2026-08-25 - Category: safety-policy-update - Coverage: 5 outlets - Announcement: yes - Group: covered - Source: https://openai.com/index/disrupting-malicious-uses-of-ai-influence-campaign-russia - Record: https://forck.live/items/4393-disrupting-a-new-covert-influence-campaign-from-russia - Subject: GPT / ChatGPT / API OpenAI enforced its safety policy by banning accounts originating from Russia that were leveraging AI to promote a fabricated think tank and a sovereignty index that favored Russia and disparaged the West. ### Hugging Face — Wire It, Run It, Deploy It: AI Workflows in Gradio - Company: Hugging Face (huggingface.co) - Announced: 2026-08-25 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://huggingface.co/blog/gradio-workflow-guide - Record: https://forck.live/items/4392-wire-it-run-it-deploy-it-ai-workflows-in-gradio - Subject: Platform Gradio introduces gr.Workflow, a new feature that allows users to build AI pipelines as a graph of typed nodes, which serves as a drag-and-drop interface, a REST API, and can be deployed to Hugging Face Spaces with one command. ### Google Antigravity — Improving the Version Control Experience - Company: Google Antigravity (antigravity.google) - Announced: 2026-08-24T20:30:00+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://antigravity.google/blog/vcs-and-terminal - Record: https://forck.live/items/4541-improving-the-version-control-experience - Subject: Antigravity The Antigravity 2.0 coding assistant now includes a VCS side panel and an embedded terminal, allowing users to perform Git operations, view diffs, stage and commit changes, and run command line tools without switching applications. This update changes the capabilities of the existing product by adding version control and terminal features directly into the agent sidebar and right panel. ### Amazon — Introducing new Ray capabilities on SageMaker HyperPod - Company: Amazon (amazon.com) - Announced: 2026-08-24T19:32:14+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://aws.amazon.com/blogs/machine-learning/introducing-new-ray-capabilities-on-sagemaker-hyperpod/ - Record: https://forck.live/items/4384-introducing-new-ray-capabilities-on-sagemaker-hyperpod - Subject: Bedrock / Nova Amazon announces new Ray capabilities on SageMaker HyperPod, integrating Ray with HyperPod's purpose-built infrastructure for foundation model training and serving. Data scientists can create Ray clusters, open Ray Dashboard and Grafana dashboards, connect JupyterLab or Code Editor workspaces, submit distributed jobs, and configure hung job detection from SageMaker Studio. The integration provides automatic fault tolerance, tiered checkpointing, and SageMaker JumpStart integration for loading model weights directly into Ray Serve endpoints. ### Amazon — Democratizing institutional knowledge: Building an AI-powered knowledge management system with AWS - Company: Amazon (amazon.com) - Announced: 2026-08-24T18:59:15+00:00 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://aws.amazon.com/blogs/machine-learning/democratizing-institutional-knowledge-building-an-ai-powered-knowledge-management-system-with-aws/ - Record: https://forck.live/items/4382-democratizing-institutional-knowledge-building-an-ai-powered-knowledge - Subject: Bedrock / Nova A customizable, smart-caching cloud-based solution for knowledge management using AWS services, including Amazon Bedrock Knowledge Bases, Amazon Titan Text Embeddings, and an avatar system. It is a how-to guide for building an AI-powered knowledge management system, not a product or model announcement. ### Meta — MetaRoCE: A New RDMA Transport Built for AI-Scale Ethernet - Company: Meta (meta.com) - Announced: 2026-08-24T18:02:29+00:00 - Category: infrastructure-release - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://engineering.fb.com/2026/08/24/networking-traffic/metaroce-rdma-transport-ai-ethernet/ - Record: https://forck.live/items/4377-metaroce-a-new-rdma-transport-built-for-ai-scale-ethernet - Subject: Llama / infrastructure Meta announced MetaRoCE, a new RDMA transport protocol designed for AI workloads on commodity Ethernet, and is releasing its specification, reference software implementation, and compliance test suite through the Open Compute Project. ### Meta — MTIA 300: Meta’s First Training Chip with Built-in NICs and Communication-Offloading Engines - Company: Meta (meta.com) - Announced: 2026-08-24T17:45:52+00:00 - Category: infrastructure-release - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://engineering.fb.com/2026/08/24/networking-traffic/mtia-300-meta-training-chip-built-in-nics/ - Record: https://forck.live/items/4376-mtia-300-meta-s-first-training-chip-with-built-in-nics-and-communication - Subject: Llama / infrastructure Meta announces MTIA 300, its first training chip with built-in NIC chiplets and communication-offloading engines, designed to optimize communication for training recommendation models, achieving 3.9x faster communication than equivalent GPU clusters on a 150-billion-parameter production model. ### Databricks — How Databricks Uses AI to Accelerate Incident Investigation - Company: Databricks (databricks.com) - Announced: 2026-08-24T17:00:00+00:00 - Category: capability-change - Coverage: not counted - Announcement: no - Group: routine - Source: https://www.databricks.com/blog/how-databricks-uses-ai-accelerate-incident-investigation - Record: https://forck.live/items/17497-how-databricks-uses-ai-to-accelerate-incident-investigation - Subject: Mosaic AI Databricks describes AI SRE, an AI-powered debugging agent that automatically triages incidents across 100s of microservices deployed on 1,500+ Kubernetes clusters. The system runs three parallel investigation tracks—platform health checks, service-level analysis, and runbook execution—to produce an initial assessment before an engineer intervenes. It is designed to reduce the 60–80% of investigation time spent on context assembly and to make expert knowledge available across 150+ teams. ### Amazon — Agentic Resource Discovery (ARD): An open specification for agent discovery - Company: Amazon (amazon.com) - Announced: 2026-08-24T16:22:03+00:00 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://aws.amazon.com/blogs/machine-learning/agentic-resource-discovery-ard-an-open-specification-for-agent-discovery/ - Record: https://forck.live/items/4374-agentic-resource-discovery-ard-an-open-specification-for-agent-discovery - Subject: Bedrock / Nova Amazon announces the Agentic Resource Discovery (ARD) open specification for federated agent discovery across environments, and the AWS Agent Registry, a centralized catalog for agents, MCP servers, and tools, now in preview. ### Amazon — Building a restaurant telephony AI host with Amazon Connect - Company: Amazon (amazon.com) - Announced: 2026-08-24T16:13:09+00:00 - Category: developer-tool-release - Coverage: not counted - Announcement: no - Group: routine - Source: https://aws.amazon.com/blogs/machine-learning/building-a-restaurant-telephony-ai-host-with-amazon-connect/ - Record: https://forck.live/items/4372-building-a-restaurant-telephony-ai-host-with-amazon-connect - Subject: Bedrock / Nova Building a restaurant telephony AI host using Amazon Connect, Amazon Lex V2, Amazon Connect Agentic Voice, Amazon Connect Customer AI agents powered by Anthropic Claude Haiku 4.5, and the Model Context Protocol (MCP). The solution includes AWS CDK deployment, AgentCore Gateway, and a restaurant backend with DynamoDB, Lambda, and API Gateway, designed to handle inbound phone calls and order fulfillment without an app or website. ### Amazon — AI-powered metadata correction and harmonization - Company: Amazon (amazon.com) - Announced: 2026-08-24T15:53:03+00:00 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://aws.amazon.com/blogs/machine-learning/ai-powered-metadata-correction-and-harmonization/ - Record: https://forck.live/items/4373-ai-powered-metadata-correction-and-harmonization - Subject: Bedrock / Nova An AI-powered metadata correction and harmonization workflow built on AWS, using Amazon Bedrock for LLM-powered schema alignment and correction recommendations. It describes a human-in-the-loop approach and discusses implementation and governance considerations. ### NVIDIA — How XPUs Meet a World-Class AI Factory - Company: NVIDIA (nvidia.com) - Announced: 2026-08-24T15:00:54+00:00 - Category: infrastructure-release - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://blogs.nvidia.com/blog/nvlink-fusion-xpu-ai-factory/ - Record: https://forck.live/items/4368-how-xpus-meet-a-world-class-ai-factory - Subject: AI platform NVIDIA announced NVLink Fusion, an infrastructure solution that connects custom XPUs to NVIDIA's NVLink scale-up domain and AI factory ecosystem, enabling faster time to market and reduced risk for semi-custom AI factories. ### NVIDIA — With Groq 3 LPX in Full Production, NVIDIA Extends Vera Rubin Inference for Agents - Company: NVIDIA (nvidia.com) - Announced: 2026-08-24T15:00:41+00:00 - Category: product-launch - Coverage: 9 outlets - Announcement: yes - Group: covered - Source: https://blogs.nvidia.com/blog/vera-rubin-lpx-spectrum-x-nvlink-fusion/ - Record: https://forck.live/items/4369-with-groq-3-lpx-in-full-production-nvidia-extends-vera-rubin-inference-for - Subject: AI platform NVIDIA announced that its Groq 3 LPX inference accelerator is in full production, delivering 3,400 output tokens per second on Gemma 4 31B for 100,000-token long-context use cases, 4x faster than alternatives. The system is part of the Vera Rubin platform and is being adopted by partners like SpaceXAI, CoreWeave, and Nebius. ### NVIDIA — Up to 30x More Work Per Watt: NVIDIA Vera Rubin NVL72 Sets a New Efficiency Standard for AI Agents - Company: NVIDIA (nvidia.com) - Announced: 2026-08-24T15:00:19+00:00 - Category: new-model - Coverage: not counted - Announcement: yes - Group: models - Source: https://blogs.nvidia.com/blog/vera-rubin-nvl72-efficiency-ai-agents/ - Record: https://forck.live/items/4370-up-to-30x-more-work-per-watt-nvidia-vera-rubin-nvl72-sets-a-new-efficiency - Subject: AI platform NVIDIA announces measured performance data for Vera Rubin NVL72 systems, showing up to 30x higher throughput per megawatt than GB300 NVL72 on agentic workloads and up to 35x lower cost per million tokens, featuring extreme codesign with disaggregated serving, expert parallelism, distributed KV-caching, and new Tensor Cores and Transformer Engine for inference efficiency on models like DeepSeek V4 Pro, Kimi K3, MiniMax M3, GLM5.3, and Qwen3.5, though it is a system architecture announcement rather than a product launch or model release, so other is most appropriate given no new model or product is introduced; however, I note the source describes a platform that enables models, so I choose 'new_model' as it is the closest fit for the Vera Rubin NVL72 architecture being new, but strictly it is neither a model nor a product launch—thus 'other' would be more accurate; reconsidering, the announcement is about a new system architecture for running models, not a model itself, so I will use 'other'—but the instructions say 'new_model' for a model, so since this is not a model, I must use 'other'. ### ElevenLabs — Introducing the ElevenLabs CLI v1 - Company: ElevenLabs (elevenlabs.io) - Announced: 2026-08-24T12:00:00+00:00 - Category: developer-tool-release - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://elevenlabs.io/blog/elevenlabs-cli-v1 - Record: https://forck.live/items/7230-introducing-the-elevenlabs-cli-v1 - Subject: Voice AI ElevenLabs announced the v1 release of its CLI, bringing API endpoints into the terminal with agent-first design, dry-run mode, and agents-as-code workflow for ElevenAgents. ### OpenAI — Advancing price-performance for developers with GPT‑5.6 in Kiro - Company: OpenAI (openai.com) - Announced: 2026-08-24T12:00:00+00:00 - Category: new-model - Coverage: not counted - Announcement: yes - Group: models - Source: https://openai.com/index/gpt-5-6-in-kiro - Record: https://forck.live/items/4385-advancing-price-performance-for-developers-with-gpt-5-6-in-kiro - Subject: GPT / ChatGPT / API GPT-5.6 is now available in Kiro, offering better price-performance for developers to plan, build, review, and test software. ### Supabase — Enterprise-managed auth for the Supabase MCP server - Company: Supabase (supabase.com) - Announced: 2026-08-24 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://supabase.com/blog/enterprise-managed-auth-for-the-supabase-mcp-server - Record: https://forck.live/items/21271-enterprise-managed-auth-for-the-supabase-mcp-server - Subject: Supabase Supabase made enterprise-managed authentication for its MCP server generally available on Team and Enterprise plans. The feature, built with Anthropic and Okta, lets IT admins control access to Supabase in Claude through their identity provider, with permissions tied to each employee's existing Supabase role. ### Thinking Machines Lab — Announcing Safety Research Grants - Company: Thinking Machines Lab (thinkingmachines.ai) - Announced: 2026-08-24 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://thinkingmachines.ai/news/safety-research-grants - Record: https://forck.live/items/18554-announcing-safety-research-grants - Subject: Thinking Machines / Inkling Thinking Machines Lab announced Tinker grants of up to $50,000 in credits to support safety research on open-weight models. The grants are intended for projects exploring differential acceleration of defensive over offensive capabilities, building classifiers for hazardous data at scale, tamper-resistant safety training, and understanding alignment failure modes such as emergent misalignment and reward hacking. ### Factory AI — Why model routing must be in the harness - Company: Factory AI (factory.com) - Announced: 2026-08-24 - Category: research-paper - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://factory.com/news/model-routing-belongs-in-the-harness - Record: https://forck.live/items/17861-why-model-routing-must-be-in-the-harness - Subject: Droid Factory AI argues that model routing should be performed inside the agent harness rather than at the API gateway, because the harness has access to session history, cache state, and task outcomes. The company's Factory Router, deployed in production for over two months, reduced aggregate inference cost by 58% and median turn latency from 81 to 49 seconds without lowering quality across eight measures. ### Liquid AI — Introducing Pipette: A benchmarking suite for on-device intelligence - Company: Liquid AI (liquid.ai) - Announced: 2026-08-24 - Category: developer-tool-release - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://www.liquid.ai/blog/pipette-on-device-ai-benchmarking-by-liquid-ai - Record: https://forck.live/items/16770-introducing-pipette-a-benchmarking-suite-for-on-device-intelligence - Subject: LFM / d1 models Liquid AI, with Artificial Analysis, released Pipette, an open-source platform for benchmarking foundation models on edge devices. The release includes a public dataset of lab-verified results covering more than 1,000 model × quantization × runtime × device × context configurations, open-source benchmark clients for macOS, Windows, iOS, and Android, and an interactive dashboard for analyzing deployment configurations. ### Anthropic — How an Anthropic field marketer uses Claude Code to send weekly personalized updates to every sales rep - Company: Anthropic (anthropic.com) - Announced: 2026-08-24 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://claude.com/blog/how-an-anthropic-field-marketer-uses-claude-code-to-send-weekly-personalized-updates-to-every-sales-rep - Record: https://forck.live/items/7418-how-an-anthropic-field-marketer-uses-claude-code-to-send-weekly-personalized - Subject: Claude An Anthropic field marketer describes using Claude Code to create a weekly personalized Slack digest for each sales rep, pulling data from BigQuery and the CRM to match marketing content to each rep's accounts and territory. ### Mistral AI — Mistral x HUMAIN - Company: Mistral AI (mistral.ai) - Announced: 2026-08-24 - Category: partnership-acquisition - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://mistral.ai/news/mistral-x-humain - Record: https://forck.live/items/4380-mistral-x-humain - Subject: Mistral / Le Chat Mistral AI and HUMAIN announced a strategic collaboration to advance sovereign AI in Saudi Arabia and the Middle East, focusing on AI infrastructure, model development, and local deployment in areas such as cybersecurity, voice, and Arabic-language frontier models. The partnership involves hundreds of millions of Euros, potential use of HUMAIN's data centers, and a joint go-to-market strategy for regulated industries.