# forck.live — 15–21 June 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-W25, 15–21 June 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-W25
- The current issue, and the catalogue of the other weeks: https://forck.live/briefing.md
- Catalogue: https://forck.live/llms.txt

## Announcements

### OpenAI — Samsung Electronics brings ChatGPT and Codex to employees

- Company: OpenAI (openai.com)
- Announced: 2026-06-21T23:00:00+00:00
- Category: availability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://openai.com/index/samsung-electronics-chatgpt-codex-deployment
- Record: https://forck.live/items/60-samsung-electronics-brings-chatgpt-and-codex-to-employees
- Subject: GPT / ChatGPT / API

Samsung Electronics is deploying ChatGPT Enterprise and Codex to its employees globally, which is one of OpenAI's largest enterprise AI rollouts.

### Sakana AI — Sakana Fugu: One Model to Command Them All

- Company: Sakana AI (sakana.ai)
- Announced: 2026-06-21T15:00:00+00:00
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://sakana.ai/fugu-release
- Record: https://forck.live/items/4681-sakana-fugu-one-model-to-command-them-all
- Subject: Sakana models

Sakana AI announces the launch of Sakana Fugu, a new product that is a multi-agent orchestration system accessible via a single API. It comes in two models: Fugu and Fugu Ultra. Fugu Ultra matches frontier performance on benchmarks without export control risks.

### Vercel — The Agent Stack

- Company: Vercel (vercel.com)
- Announced: 2026-06-20T04:30:00+00:00
- Category: developer-tool-release
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://vercel.com/blog/agent-stack
- Record: https://forck.live/items/20698-the-agent-stack
- Subject: v0 / AI SDK / AI Gateway

Vercel introduced the Agent Stack, a set of building blocks for creating and shipping production-grade agents. The stack includes AI SDK for a unified model interface, AI Gateway for routing and cost tracking, Workflow SDK for durable workflows, Vercel Sandbox for isolated code execution, and Vercel Connect for secure data and tool access.

### TypeSafe AI — AI: too good to be true, too bad to be useful

- Company: TypeSafe AI (typesafe.ai)
- Announced: 2026-06-19
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://typesafe.ai/blog/ai-too-good-to-be-true-too-bad-to-be-useful-typesafe-ai
- Record: https://forck.live/items/12214-ai-too-good-to-be-true-too-bad-to-be-useful
- Subject: Jev / System One

TypeSafe AI argues that RLHF-trained language models are designed to please humans and assist, not to make reliable autonomous decisions, and asks what comes next.

### Amp — Custom Agents

- Company: Amp (ampcode.com)
- Announced: 2026-06-19
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://ampcode.com/news/custom-agents
- Record: https://forck.live/items/7735-custom-agents
- Subject: Amp

Amp announced that users can now create custom agents with plugins, which can be used as main agents, subagents, or in tool pipelines. Custom agents can have custom colors and are defined via Amp plugins.

### Hugging Face — MosaicLeaks: Can your research agent keep a secret?

- Company: Hugging Face (huggingface.co)
- Announced: 2026-06-18T18:13:13+00:00
- Category: research-paper
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://huggingface.co/blog/ServiceNow/mosaicleaks
- Record: https://forck.live/items/1479-mosaicleaks-can-your-research-agent-keep-a-secret
- Subject: Platform

Introduces MosaicLeaks, a benchmark for evaluating privacy leakage in deep research agents, and proposes Privacy-Aware Deep Research (PA-DR) training that reduces answer/full-information leakage from 34.0% to 9.9% while improving strict chain success from 48.7% to 58.7%.

### OpenAI — New usage analytics and updated spend controls for enterprises

- Company: OpenAI (openai.com)
- Announced: 2026-06-18T17:00:00+00:00
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://openai.com/index/chatgpt-enterprise-spend-controls
- Record: https://forck.live/items/61-new-usage-analytics-and-updated-spend-controls-for-enterprises
- Subject: GPT / ChatGPT / API

OpenAI introduces new spend controls and usage analytics for ChatGPT Enterprise, helping organizations manage costs and scale AI with confidence.

### Black Forest Labs — Our co-founder and CEO urges G7 leaders to back open innovation

- Company: Black Forest Labs (bfl.ai)
- Announced: 2026-06-18T16:55:00+00:00
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://bfl.ai/blog/our-co-founder-and-ceo-urges-g7-leaders-to-back-open-innovation
- Record: https://forck.live/items/3386-our-co-founder-and-ceo-urges-g7-leaders-to-back-open-innovation
- Subject: FLUX

The co-founder and CEO of Black Forest Labs urges G7 leaders to back open innovation.

### Harvey — Debunking the Myths of Data Sovereignty and US-Based Service Providers

- Company: Harvey (harvey.ai)
- Announced: 2026-06-18T12:00:00+00:00
- Category: not stated
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://www.harvey.ai/blog/debunking-myths-data-sovereignty-and-us-based-service-providers
- Record: https://forck.live/items/17547-debunking-the-myths-of-data-sovereignty-and-us-based-service-providers
- Subject: Harvey

Our General Counsel shares how we approach government data requests and why the scenario of “secret subpoenas” is exceedingly unlikely for Harvey.

### Harvey — Visualize Legal Data on Demand in Harvey

- Company: Harvey (harvey.ai)
- Announced: 2026-06-18T12:00:00+00:00
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://www.harvey.ai/blog/visualize-legal-data-in-harvey
- Record: https://forck.live/items/17546-visualize-legal-data-on-demand-in-harvey
- Subject: Harvey

Harvey added a feature that lets users generate interactive timelines, entity charts, compliance matrices, and dashboards directly in the platform from unstructured legal inputs. The outputs can be downloaded as shareable .html files without extra tools.

### ElevenLabs — Poland invests in ElevenLabs

- Company: ElevenLabs (elevenlabs.io)
- Announced: 2026-06-18T12:00:00+00:00
- Category: funding-company-news
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://elevenlabs.io/blog/poland-invests-in-elevenlabs
- Record: https://forck.live/items/3350-poland-invests-in-elevenlabs
- Subject: Voice AI

Poland has made an investment in ElevenLabs.

### OpenAI — Improving health intelligence in ChatGPT

- Company: OpenAI (openai.com)
- Announced: 2026-06-18T11:00:00+00:00
- Category: model-update
- Coverage: not counted
- Announcement: yes
- Group: models
- Source: https://openai.com/index/improving-health-intelligence-in-chatgpt
- Record: https://forck.live/items/62-improving-health-intelligence-in-chatgpt
- Subject: GPT / ChatGPT / API

GPT-5.5 Instant improves ChatGPT's health and wellness responses with stronger reasoning, better context, clearer communication, and physician-informed evaluations.

### OpenAI — Using AI to help physicians diagnose rare genetic diseases affecting children

- Company: OpenAI (openai.com)
- Announced: 2026-06-18T08:00:00+00:00
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://openai.com/index/diagnose-rare-childhood-diseases
- Record: https://forck.live/items/63-using-ai-to-help-physicians-diagnose-rare-genetic-diseases-affecting-children
- Subject: GPT / ChatGPT / API

Researchers used an OpenAI reasoning model to help diagnose rare diseases, identifying 18 new diagnoses in previously unsolved cases.

### Ai2 — How Domyn and AISquared built on Ai2's open releases

- Company: Ai2 (allenai.org)
- Announced: 2026-06-18
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://allenai.org/blog/domyn-aisquared-testimonial
- Record: https://forck.live/items/18550-how-domyn-and-aisquared-built-on-ai2-s-open-releases
- Subject: Olmo / Molmo / Asta

Domyn and AISquared, two AI labs serving regulated industries, built models on Ai2's open releases. AISquared's Bolt Instruct family is fine-tuned from Olmo 2, Olmo 3, and Olmo 3.1, and migrating to Bolt cut AISquared's infrastructure hosting costs by roughly 50%. Domyn's Domyn Small, a 10B-parameter reasoning model, used Ai2's Dolma and Dolci datasets, gaining 10.1 points on GPQA-Diamond from Dolci.

### Kyutai — The FID Lottery

- Company: Kyutai (kyutai.org)
- Announced: 2026-06-18
- Category: research-paper
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://kyutai.org/fid-lottery
- Record: https://forck.live/items/18339-the-fid-lottery
- Subject: Moshi / Pocket TTS

Kyutai's research shows that reported FID scores are heavily influenced by random seeds in training and generation, with training seed variance dominating. The paper finds that a single-seed FID sits on a noise floor that recipe-level gains regularly fall below, and that a lucky training seed can reach the same FID with up to 2× less compute than an unlucky one.

### Cua — Inside Linux computer-use: AT-SPI, XTEST, and background agents

- Company: Cua (cua.ai)
- Announced: 2026-06-18
- Category: developer-tool-release
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://cua.ai/blog/inside-linux-computer-use
- Record: https://forck.live/items/16820-inside-linux-computer-use-at-spi-xtest-and-background-agents
- Subject: Cua Driver / Lume / Cua Bench / Cua Fleets

Cua Driver 0.5.7 adds background computer-use for Linux desktop apps, using AT-SPI 2 over D-Bus for the accessibility tree and XTEST for input synthesis, with a painted agent cursor kept separate from the physical pointer. The supported backend is X11 and XWayland, with native Wayland in preview behind an opt-in flag. It is available today for any agent over MCP or the CLI, tested on Debian 12, Ubuntu 22.04, Ubuntu 24.04, Rocky 9, and Fedora 41.

### Liquid AI — LFM2.5 Retrievers: Bi-directional LFMs for Fast Multilingual Search

- Company: Liquid AI (liquid.ai)
- Announced: 2026-06-18
- Category: new-model
- Coverage: not counted
- Announcement: yes
- Group: models
- Source: https://www.liquid.ai/blog/lfm2-5-retrievers
- Record: https://forck.live/items/16782-lfm2-5-retrievers-bi-directional-lfms-for-fast-multilingual-search
- Subject: LFM / d1 models

Liquid AI released two 350M-parameter multilingual retrieval models, LFM2.5-ColBERT-350M and LFM2.5-Embedding-350M, the first bidirectional members of the LFM family. They support multilingual and cross-lingual search across 11 languages and are designed for short-context search tasks such as product catalogs and FAQ knowledge bases. The embedding model produces a single vector per document for fast search, while the ColBERT model uses per-token vectors for higher accuracy.

### xAI — Introducing Grok for Word

- Company: xAI (x.ai)
- Announced: 2026-06-18
- Category: not stated
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://x.ai/news/introducing-word-addin
- Record: https://forck.live/items/16460-introducing-grok-for-word
- Subject: Grok

Grok for Word Use the Grok add-in for Microsoft Word to turn notes into documents, style and format your work, or bring research from the web into Word. Jun 18, 2026

### Ideogram — How to JSON prompt for Ideogram 4.0

- Company: Ideogram (ideogram.ai)
- Announced: 2026-06-18
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://ideogram.ai/blog/ideogram-4-json-prompting
- Record: https://forck.live/items/15462-how-to-json-prompt-for-ideogram-4-0
- Subject: Ideogram image models

Ideogram explains that version 4.0 was trained on structured JSON captions, making the model more literal and responsive to specific, labeled descriptions rather than plain-text prompts. The guide covers how to write JSON prompts, the importance of style_description fields, and the built-in Magic Prompt translator that converts natural language into structured captions.

### Manus — How I Turned a Course Page Into a Course Business With Website Builder

- Company: Manus (manus.im)
- Announced: 2026-06-18
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://manus.im/blog/build-a-course-business-with-manus-web-builder
- Record: https://forck.live/items/14711-how-i-turned-a-course-page-into-a-course-business-with-website-builder
- Subject: Manus

Manus published a walkthrough showing how its Website Builder can create a full course business—including payments, protected lessons, AI-driven practice, scheduled follow-ups, and traffic analytics—within a single project. The article describes the builder's capabilities for connecting Stripe checkout, role-based access, custom domains, built-in AI features, automated schedules, and analytics. It notes that built-in analytics cover traffic-level metrics only (page views, visits, visitors, average duration, bounce rate, viewed pages, referrers, regions, devices) and that product-specific reporting requires custom logic.

### Amp — A Faster Librarian

- Company: Amp (ampcode.com)
- Announced: 2026-06-18
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://ampcode.com/news/a-faster-librarian
- Record: https://forck.live/items/7736-a-faster-librarian
- Subject: Amp

Amp announced that The Librarian is now approximately 3 times faster and 43% cheaper while maintaining the same quality. The update uses GPT-5.5 with websocket mode and an updated system prompt, enabling more parallel tool calls and fewer turns per search.

### Anthropic — Claude Code now supports artifacts

- Company: Anthropic (anthropic.com)
- Announced: 2026-06-18
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://claude.com/blog/artifacts-in-claude-code
- Record: https://forck.live/items/7403-claude-code-now-supports-artifacts
- Subject: Claude

Anthropic announced that Claude Code now supports artifacts, which capture work progress as live, shareable visual pages that update in place. Artifacts are private to the organization by default and are available in beta to Claude Team and Enterprise orgs.

### xAI — Grok for Word

- Company: xAI (x.ai)
- Announced: 2026-06-18
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://x.ai/news/introducing-word-addin
- Record: https://forck.live/items/3449-grok-for-word
- Subject: Grok

xAI introduced a Grok add-in for Microsoft Word, allowing users to convert notes into documents, style and format text, and integrate web research into Word.

### xAI — Grok on Databricks

- Company: xAI (x.ai)
- Announced: 2026-06-18
- Category: availability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://x.ai/news/grok-databricks
- Record: https://forck.live/items/3448-grok-on-databricks
- Subject: Grok

Grok models are now available on Databricks Agent Bricks.

### Cursor — Improvements to Cursor Automations

- Company: Cursor (cursor.com)
- Announced: 2026-06-18
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://cursor.com/changelog/06-18-26
- Record: https://forck.live/items/2596-improvements-to-cursor-automations
- Subject: Cursor

Cursor adds the /automate skill, new GitHub and Slack triggers, computer use support, and several improvements to its Automations feature.

### Hugging Face — Is it agentic enough? Benchmarking open models on your own tooling

- Company: Hugging Face (huggingface.co)
- Announced: 2026-06-18
- Category: research-paper
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://huggingface.co/blog/is-it-agentic-enough
- Record: https://forck.live/items/1481-is-it-agentic-enough-benchmarking-open-models-on-your-own-tooling
- Subject: Platform

Hugging Face introduces a benchmark and harness for evaluating how efficiently open models use the transformers library in agentic tasks, measuring not just correctness but also cost, latency, and token usage across different library configurations.

### Hugging Face — Beyond LoRA: Can you beat the most popular fine-tuning technique?

- Company: Hugging Face (huggingface.co)
- Announced: 2026-06-18
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://huggingface.co/blog/peft-beyond-lora
- Record: https://forck.live/items/1480-beyond-lora-can-you-beat-the-most-popular-fine-tuning-technique
- Subject: Platform

Parameter-efficient fine-tuning (PEFT) techniques, questioning whether LoRA is always the best choice, and introduces the PEFT library and benchmarks to help users make informed decisions.

### Replit — Replit is now available in Claude

- Company: Replit (replit.com)
- Announced: 2026-06-17T19:05:46+00:00
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://replit.com/blog/replit-claude
- Record: https://forck.live/items/17894-replit-is-now-available-in-claude
- Subject: Replit Agent

Replit announced its integration directly inside Claude, allowing users to design apps in Claude and send them to Replit for building and shipping without context switching. Claude can also delegate general development tasks to Replit via the official Replit Connector.

### Swiss AI Initiative — SME Circle #4

- Company: Swiss AI Initiative (swiss-ai.org)
- Announced: 2026-06-17T19:00:00+00:00
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://apertus-ai.org/articles/2026-06-sme-circle/
- Record: https://forck.live/items/18375-sme-circle-4
- Subject: Apertus

The Swiss AI Initiative's SME Circle #4 event at the ETH AI Center gathered around seventy participants to discuss open-source AI adoption. The initiative announced the upload of Apertus 1.1 “Mini” with three compact sizes for offline use, and previewed the upcoming Apertus 1.5 release featuring multimodal input and improved instruction calling. A new open access platform, Apertus Claritas, was also announced for interpretability tech notes.

### Black Forest Labs — How Envato Built Its Creative AI Engine on FLUX

- Company: Black Forest Labs (bfl.ai)
- Announced: 2026-06-17T14:29:00+00:00
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://bfl.ai/blog/how-envato-built-its-creative-ai-engine-on-flux
- Record: https://forck.live/items/3387-how-envato-built-its-creative-ai-engine-on-flux
- Subject: FLUX

Envato built its creative AI engine on FLUX.

### Amazon — From the Hugging Face Hub to robot hardware with Strands Agents and LeRobot

- Company: Amazon (amazon.com)
- Announced: 2026-06-17T10:18:05+00:00
- Category: developer-tool-release
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://huggingface.co/blog/amazon/strands-lerobot-hub-to-hardware
- Record: https://forck.live/items/1482-from-the-hugging-face-hub-to-robot-hardware-with-strands-agents-and-lerobot

AWS announces the Strands Robots open-source SDK, integrating LeRobot for robot learning from simulation to hardware, allowing a single agent loop to record demonstrations, run policies, and deploy to physical robots with a one-argument change from sim to real mode, and supporting fleet coordination via a peer mesh and Zenoh broadcast, with integration to the Hugging Face Hub for dataset sharing and policy checkpoint inference via GR00T and MolmoAct2 checkpoints through the LerobotLocal path, all under Apache 2.0 license, with the runnable example available on GitHub that works in simulation without hardware, GPU, or Hugging Face credentials, and the published date is June 17, 2026 in the source text but this appears to be a future date, which we note based on the source text's claim of 'Published June 17, 2026' without any indication of error, but we must follow the rule to not use prior knowledge about the date's validity, so we report it as stated in the source text, and we also note that the source text is a blog post from Hugging Face authored by Sundar Raghavan and Cagatay Cali, both affiliated with Amazon, and the post is classified as Enterprise Article, and the walkthrough covers five steps inside a single agent: build the agent, record a demonstration as a LeRobotDataset in simulation, run a policy, deploy the same agent code to a physical SO-101 robot with one keyword argument change, and broadcast commands across a fleet over the Zenoh mesh, with the dataset format shared between sim and hardware, and the policy providers including GR00T and LerobotLocal, and the integration with the Hugging Face Hub for pushing datasets and pulling policy checkpoints, and the minimal prerequisites are Python 3.12+, Linux or macOS, a Strands-compatible model provider for reasoning, and Strands Robots installed with the [sim-mujoco,lerobot,mesh] extras, and the advanced prerequisites include a Hugging Face account with write permission, an SO-101 follower and leader pair, calibration files, and for local GR00T inference an NVIDIA GPU with at least 16 GB of video memory and Docker, and the example runs end-to-end on a laptop with the minimal setup, and the notebook is the recommended starting point, and the source text also mentions the LeRobot integration in Strands Robots, and the dataset format stays exactly as LeRobot wrote it, and the agent loop is the glue, and the design choices are Robot("so100") defaults to simulation, and mode="real" returns a hardware robot driven by LeRobot, and both modes share the same DatasetRecorder and policy providers, and the whole workflow in five lines of Python is shown, and the source text also includes a figure description of Robot("so100") defaults to MuJoCo-backed simulation, and mode="real" returns a hardware robot driven by LeRobot, and the DatasetRecorder is shared between simulation and hardware recording, and the step-by-step instructions for setup, recording demonstrations, and pushing to the Hub are provided, but the source text is truncated after Step 2, so we cannot extract further details about the remaining steps, but we note that the source text is a blog post from Hugging Face, and the announcement is about the Strands Robots SDK, which is an open source SDK from AWS, and the integration with LeRobot, and the support for GR00T and MolmoAct2 checkpoints, and the license is Apache 2.0, and the date is June 17, 2026, and the models affected are GR00T and MolmoAct2, and the category is developer_tool_release, and the evidence excerpt is the verbatim span that states the SDK is from AWS under Apache 2.0, and the summary is as above, and the additional_notes could include the future date observation, but we leave it empty as per the instruction to not introduce prior knowledge about the date's validity, and we note that the source text does not state any context_window or pricing_note, so those are left empty, and the models_affected list includes only GR00T and MolmoAct2 as they appear in the source text, and the license is Apache 2.0, and the provider is AWS, and the title is as given, and the url is as given, and the date is 2026-06-17, and the category is developer_tool_release, and the summary is a concise description of the announcement, and the evidence_excerpt supports the summary, and the JSON is structured as per the schema, and we ensure no fields are missing, and we use the exact values from the source text, and we do not include any extra information, and we follow all rules, and we output the JSON in the required format.

### OpenAI — A near-autonomous AI chemist improves a challenging reaction in medicinal chemistry

- Company: OpenAI (openai.com)
- Announced: 2026-06-17T10:00:00+00:00
- Category: research-paper
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://openai.com/index/ai-chemist-improves-reaction
- Record: https://forck.live/items/64-a-near-autonomous-ai-chemist-improves-a-challenging-reaction-in-medicinal
- Subject: GPT / ChatGPT / API

OpenAI and Molecule.one demonstrated a near-autonomous AI chemist using GPT-5.4 that improved a key drug-making reaction, advancing medicinal chemistry research.

### Z.ai — GLM-5.2: Built for Long-Horizon Tasks

- Company: Z.ai (z.ai)
- Announced: 2026-06-17T09:01:25+00:00
- Category: new-model
- Coverage: 1 outlet
- Announcement: yes
- Group: models
- Source: https://huggingface.co/blog/zai-org/glm-52-blog
- Record: https://forck.live/items/1483-glm-5-2-built-for-long-horizon-tasks

Z.AI introduces GLM-5.2, a model for long-horizon tasks with a solid 1M-token context, improved architecture using IndexShare, MIT open-source license, and strong benchmark performance including 81.0 on Terminal-Bench 2.1 and 62.1 on SWE-bench Pro, outperforming its predecessor GLM-5.1 and closing the gap to closed-source frontier models like Opus 4.8 and GPT-5.5 on several benchmarks.

### Databricks — What’s New in the AI Platform: Agents for ML Engineering, Our Deep Learning Platform, and New Capabilities for Real-Time ML

- Company: Databricks (databricks.com)
- Announced: 2026-06-17T08:44:14+00:00
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://www.databricks.com/blog/whats-new-ai-platform-agents-ml-engineering-our-deep-learning-platform-and-new-capabilities
- Record: https://forck.live/items/17510-what-s-new-in-the-ai-platform-agents-for-ml-engineering-our-deep-learning
- Subject: Mosaic AI

Databricks announced Genie Code for ML, an agent that assists with the full ML lifecycle; AI Runtime, a serverless GPU training environment now in public preview; and enhanced real-time ML support including low-latency, high-QPS serving for Feature Store and Model Serving.

### Vercel — Introducing eve

- Company: Vercel (vercel.com)
- Announced: 2026-06-17T04:00:00+00:00
- Category: developer-tool-release
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://vercel.com/blog/introducing-eve
- Record: https://forck.live/items/20700-introducing-eve
- Subject: v0 / AI SDK / AI Gateway

Vercel introduced eve, an open-source agent framework for building, running, and scaling agents. The framework includes built-in production features such as durable execution, sandboxed compute, human-in-the-loop approvals, subagents, and evals. eve is designed so that each agent is defined as a directory of files, and it supports connections to services like Slack, GitHub, Snowflake, Salesforce, Notion, and Linear via MCP servers or OpenAPI documents.

### Vercel — Introducing Vercel Connect

- Company: Vercel (vercel.com)
- Announced: 2026-06-17T04:00:00+00:00
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://vercel.com/blog/introducing-vercel-connect
- Record: https://forck.live/items/20699-introducing-vercel-connect
- Subject: v0 / AI SDK / AI Gateway

Vercel introduced Vercel Connect, a service that replaces long-lived provider tokens with runtime credential exchange for agents. Connectors are registered once and attached to projects and environments, providing short-lived, scoped tokens via OIDC identity verification. The service is in Public Beta as of August 25, 2026.

### Odyssey — Our $310 Million Fundraise to Accelerate World Simulation

- Company: Odyssey (odyssey.systems)
- Announced: 2026-06-17
- Category: funding-company-news
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://odyssey.systems/our-series-b
- Record: https://forck.live/items/18593-our-310-million-fundraise-to-accelerate-world-simulation
- Subject: Odyssey world models

Odyssey announced a $310 million Series B funding round at a $1.45 billion valuation, led by Natural Capital with participation from Amazon, GV, AMD Ventures, EQT, IQT, and others. The company stated the investment will accelerate its development of world models, and that Amazon Web Services will become its preferred cloud provider while Odyssey collaborates with Amazon's Annapurna Labs to optimize its models on AWS Trainium chips.

### Ai2 — MolmoMotion: Language-guided 3D motion forecasting

- Company: Ai2 (allenai.org)
- Announced: 2026-06-17
- Category: new-model
- Coverage: not counted
- Announcement: yes
- Group: models
- Source: https://allenai.org/blog/molmo-motion
- Record: https://forck.live/items/18551-molmomotion-language-guided-3d-motion-forecasting
- Subject: Olmo / Molmo / Asta

Ai2 released MolmoMotion, a motion forecasting model that predicts future 3D point trajectories from a video frame, query points, and a language instruction. The model uses Molmo 2 as its backbone and comes in two variants: an autoregressive version (MolmoMotion-AR) and a flow-matching version (MolmoMotion-FM). Alongside the model, Ai2 published the MolmoMotion-1M dataset of 1.16M videos with 3D point trajectories and action descriptions, and PointMotionBench, a human-validated benchmark of 2.7K video clips for evaluating object-centric 3D motion forecasting.

### Factory AI — AutoWiki

- Company: Factory AI (factory.com)
- Announced: 2026-06-17
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://factory.com/news/wiki
- Record: https://forck.live/items/17851-autowiki
- Subject: Droid

Factory released AutoWiki, a tool that automatically generates and maintains structured documentation for each repository in an organization. It analyzes the codebase to produce sections covering architecture, key flows, and component relationships, and refreshes the wiki when the repository changes. AutoWiki is available to all plans and can be accessed via a web viewer, synced to a GitHub wiki tab, used in Droid sessions, or committed directly to the repository.

### LG AI Research — LG AI Research and D&D Pharmatech Partner to Advance AI-Powered Next-Generation Peptide Drug Discovery

- Company: LG AI Research (lgresearch.ai)
- Announced: 2026-06-17
- Category: partnership-acquisition
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://www.lgresearch.ai/news/view?seq=657
- Record: https://forck.live/items/10990-lg-ai-research-and-d-d-pharmatech-partner-to-advance-ai-powered-next
- Subject: EXAONE

LG AI Research and D&D Pharmatech signed a definitive agreement to jointly develop next-generation peptide therapeutics. LG AI Research will design drug candidates using AI, while D&D Pharmatech will handle validation, clinical development, and regulatory approval. The collaboration aims to reduce development timelines and costs while improving clinical success probability.

### LG AI Research — LG AI Research and D&D Pharmatech Partner to Advance AI-Powered Next-Generation Peptide Drug Discovery

- Company: LG AI Research (lgresearch.ai)
- Announced: 2026-06-17
- Category: partnership-acquisition
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://www.lgresearch.ai/blog/view?seq=657
- Record: https://forck.live/items/4746-lg-ai-research-and-d-d-pharmatech-partner-to-advance-ai-powered-next
- Subject: EXAONE

LG AI Research and D&D Pharmatech have signed a definitive agreement to jointly develop next-generation peptide therapeutics, combining AI-driven drug discovery with oral delivery technologies.

### xAI — Grok on Amazon Bedrock

- Company: xAI (x.ai)
- Announced: 2026-06-17
- Category: availability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://x.ai/news/grok-amazon-bedrock
- Record: https://forck.live/items/3450-grok-on-amazon-bedrock
- Subject: Grok

Grok models are now available via Amazon Bedrock.

### Cohere — LLM Serving Fairness: No more noisy neighbors

- Company: Cohere (cohere.com)
- Announced: 2026-06-17
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://cohere.com/blog/serving-fairness
- Record: https://forck.live/items/3277-llm-serving-fairness-no-more-noisy-neighbors
- Subject: Command

LLM serving fairness and noisy neighbors, but the source text provides no details about a specific model, product, or announcement.

### Cursor — Cloud Environment Setup and Cloud Subagents in Agents Window

- Company: Cursor (cursor.com)
- Announced: 2026-06-17
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://cursor.com/changelog/cloud-in-agents-window
- Record: https://forck.live/items/2597-cloud-environment-setup-and-cloud-subagents-in-agents-window
- Subject: Cursor

Cursor adds cloud environment setup, cloud subagents via /in-cloud, and handoff between local and cloud to the Agents Window of the desktop app.

### Together AI — Kimi K2.7 Code vs Claude Fable 5: Landing pages that cost 94% less

- Company: Together AI (together.ai)
- Announced: 2026-06-17
- Category: research-paper
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://www.together.ai/blog/kimi-k2-7-code-vs-claude-fable-5
- Record: https://forck.live/items/2341-kimi-k2-7-code-vs-claude-fable-5-landing-pages-that-cost-94-less
- Subject: Inference platform

Together AI ran an experiment comparing Kimi K2.7 Code and Claude Fable 5 on generating 12 landing pages. Kimi cost about 94% less than Fable and produced similar-quality output, especially when given design inspiration via a custom MCP server. The results are published on the OVSC website.

### Hugging Face — Agentic Resource Discovery: Let agents search

- Company: Hugging Face (huggingface.co)
- Announced: 2026-06-17
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://huggingface.co/blog/agentic-resource-discovery-launch
- Record: https://forck.live/items/1484-agentic-resource-discovery-let-agents-search
- Subject: Platform

Hugging Face announces the Agentic Resource Discovery (ARD) specification, a draft open standard for federated discovery of tools, skills, and agents, along with a reference implementation called the Hugging Face Discover Tool.

### OpenAI — Introducing LifeSciBench

- Company: OpenAI (openai.com)
- Announced: 2026-06-17
- Category: research-paper
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://openai.com/index/introducing-life-sci-bench
- Record: https://forck.live/items/65-introducing-lifescibench
- Subject: GPT / ChatGPT / API

OpenAI introduces LifeSciBench, an expert-authored and expert-reviewed benchmark for evaluating AI systems on real-world life science research tasks and decisions.

### Vercel — Vercel for Enterprise Apps and Agents

- Company: Vercel (vercel.com)
- Announced: 2026-06-16T23:00:00+00:00
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://vercel.com/blog/vercel-for-enterprise-apps-and-agents
- Record: https://forck.live/items/20701-vercel-for-enterprise-apps-and-agents
- Subject: v0 / AI SDK / AI Gateway

Vercel introduced Enterprise Apps and Agents, a platform that adds default security and access controls for internal AI apps and agents. It includes Vercel Passport for identity-provider-based access, Vercel Connect for short-lived credentials, Enterprise Managed Users for lifecycle management, and bring-your-own-cloud on AWS. v0 now connects to Snowflake for building data apps with IDP-controlled access.

### Midjourney — Hardware announcement livestream

- Company: Midjourney (midjourney.com)
- Announced: 2026-06-16T22:08:03+00:00
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://updates.midjourney.com/hardware-announcement-livestream/
- Record: https://forck.live/items/18586-hardware-announcement-livestream
- Subject: Midjourney V8

Midjourney announced it will unveil its first secret hardware project on Wednesday, June 17, describing it as ambitious, physically large, and unexpected. The unveiling will be livestreamed at 6pm Pacific time on Discord and X.

### Google DeepMind — Unlocking UK house-building with AI-accelerated planning

- Company: Google DeepMind (deepmind.google)
- Announced: 2026-06-16T21:29:50+00:00
- Category: partnership-acquisition
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://deepmind.google/blog/unlocking-uk-house-building-with-ai-accelerated-planning/
- Record: https://forck.live/items/1211-unlocking-uk-house-building-with-ai-accelerated-planning
- Subject: Gemini

UK government and Google DeepMind partner to develop an AI prototype for accelerating housing planning decisions.

### Google Research — From pixels to planning: Earth AI for nature restoration

- Company: Google Research (research.google)
- Announced: 2026-06-16T17:30:00+00:00
- Category: research-paper
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://research.google/blog/from-pixels-to-planning-earth-ai-for-nature-restoration/
- Record: https://forck.live/items/1331-from-pixels-to-planning-earth-ai-for-nature-restoration
- Subject: Research

Google Research developed a high-resolution deep learning framework to map fine-scale ecological features like hedgerows and copses, and released a vectorized dataset (Vectorized Farmscapes 2020) for the UK to support nature restoration and carbon accounting.

### Google DeepMind — Securing the future of AI agents

- Company: Google DeepMind (deepmind.google)
- Announced: 2026-06-16T15:46:31+00:00
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://deepmind.google/blog/securing-the-future-of-ai-agents/
- Record: https://forck.live/items/1212-securing-the-future-of-ai-agents
- Subject: Gemini

Google DeepMind describes an AI Control Roadmap for securing internal systems using traditional safeguards and real-time monitoring.

### Harvey — Harvey’s Legal Intelligence Now Lives Inside Microsoft 365

- Company: Harvey (harvey.ai)
- Announced: 2026-06-16T15:30:00+00:00
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://www.harvey.ai/blog/harvey-copilot-cowork-launch
- Record: https://forck.live/items/17548-harvey-s-legal-intelligence-now-lives-inside-microsoft-365
- Subject: Harvey

Harvey integrated its legal AI as an agent inside Microsoft 365 Copilot and as a plugin inside Copilot Cowork, allowing legal professionals to ask questions, research issues, analyze documents, and retrieve content from their Harvey Vault without leaving the Microsoft 365 environment. The integration supports use cases such as contract review, M&A due diligence, opposition drafting, and document drafting, with deeper analysis available by moving threads into Harvey's web environment.

### Databricks — Sciene AI Companion: building an autonomous Customer Success platform on Databricks

- Company: Databricks (databricks.com)
- Announced: 2026-06-16T14:29:06+00:00
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://www.databricks.com/blog/sciene-ai-companion-building-autonomous-customer-success-platform-databricks
- Record: https://forck.live/items/17511-sciene-ai-companion-building-an-autonomous-customer-success-platform-on
- Subject: Mosaic AI

Sciene's AI Companion, built on Databricks, helps Quartile's Customer Success team generate context-aware emails, meeting decks, and account diagnostics. The platform reduced email reply time from 15–30 minutes to about 3 minutes, deck preparation from over 2 hours to around 10 minutes, and account diagnosis from 30+ minutes to ~5 minutes.

### Alibaba — Qwen-Robot Suite: A Foundation Model Suite for Physical World Intelligence

- Company: Alibaba (alibaba.com)
- Announced: 2026-06-16T02:00:00+00:00
- Category: new-model
- Coverage: not counted
- Announcement: yes
- Group: models
- Source: https://qwen.ai/blog?id=qwen-robotsuite
- Record: https://forck.live/items/4615-qwen-robot-suite-a-foundation-model-suite-for-physical-world-intelligence
- Subject: Qwen

Alibaba announces the Qwen-Robot Suite, a set of three foundation models (Qwen-RobotNav, Qwen-RobotManip, Qwen-RobotWorld) that bridge vision-language understanding with physical action in navigation, manipulation, and world dynamics.

### xAI — Introducing Grok for PowerPoint

- Company: xAI (x.ai)
- Announced: 2026-06-16
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://x.ai/news/introducing-powerpoint-addin
- Record: https://forck.live/items/16461-introducing-grok-for-powerpoint
- Subject: Grok

xAI made Grok available inside Microsoft PowerPoint, allowing users to turn outlines into slides, expand decks, and tighten narratives without leaving the application.

### Luma — Introducing Luma Skills: Build a Creative Workflow Once, Run It Forever

- Company: Luma (lumalabs.ai)
- Announced: 2026-06-16
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://lumalabs.ai/news/luma-skills
- Record: https://forck.live/items/7961-introducing-luma-skills-build-a-creative-workflow-once-run-it-forever
- Subject: Ray / Dream Machine

Luma introduced Skills, a repeatable workflow inside Luma Agents that allows users to build a creative process once and run it repeatedly on new assets, ensuring consistent quality. Skills consist of input, tools, and instructions, and can be built from scratch or generated by the agent. They can be shared via links, bundled into packages, or downloaded as files.

### Amp — Diffs

- Company: Amp (ampcode.com)
- Announced: 2026-06-16
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://ampcode.com/news/diffs
- Record: https://forck.live/items/7737-diffs
- Subject: Amp

Amp announced that users can now review code changes (diffs) for any thread directly within the product, on desktop or mobile. The diffing algorithm includes duplicate block detection to reduce cognitive load when reviewing large agent-generated changesets.

### Alibaba — Qwen-RobotManip: Alignment Unlocks Scale for Robotic Manipulation Foundation Models

- Company: Alibaba (alibaba.com)
- Announced: 2026-06-16
- Category: new-model
- Coverage: not counted
- Announcement: yes
- Group: models
- Source: https://qwen.ai/blog?id=qwen-robotmanip
- Record: https://forck.live/items/4614-qwen-robotmanip-alignment-unlocks-scale-for-robotic-manipulation-foundation
- Subject: Qwen

Alibaba announces Qwen-RobotManip, a generalizable Vision-Language-Action (VLA) foundation model built upon Qwen-VL, which introduces a unified alignment framework across representation, motion, and behavioral dimensions for robotic manipulation. The model uses only open-source robotic manipulation datasets and human demonstration videos to construct a ~38,100 hours pretraining corpus, and demonstrates emergent generalization capabilities across various real-robot platforms and tasks.

### Alibaba — Qwen-RobotWorld: Boundless Worlds for Embodied Agents

- Company: Alibaba (alibaba.com)
- Announced: 2026-06-16
- Category: research-paper
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://qwen.ai/blog?id=qwen-robotworld
- Record: https://forck.live/items/4613-qwen-robotworld-boundless-worlds-for-embodied-agents
- Subject: Qwen

Alibaba introduces Qwen-RobotWorld, a dual-stream diffusion world model that uses natural language as a universal action interface to unify 20+ robot embodiments and 500+ action categories, enabling cross-scenario physical generalization across manipulation, driving, and navigation. The model employs Qwen2.5-VL as an action encoder within an MMDiT architecture, and is trained on the Embodied World Knowledge (EWK) dataset of 8.6M video-text pairs. It also supports multi-view geometrically consistent generation and human-to-robot transfer via Scene2Robot.

### Alibaba — Qwen-RobotNav: A Scalable Navigation Model Designed for an Agentic Navigation System

- Company: Alibaba (alibaba.com)
- Announced: 2026-06-16
- Category: new-model
- Coverage: not counted
- Announcement: yes
- Group: models
- Source: https://qwen.ai/blog?id=qwen-robotnav
- Record: https://forck.live/items/4611-qwen-robotnav-a-scalable-navigation-model-designed-for-an-agentic-navigation
- Subject: Qwen

Alibaba presents Qwen-RobotNav, a scalable navigation model built on Qwen3-VL with a parameterised interface for controllable observation parameters. Trained on 15.6 million samples, it achieves state-of-the-art across five navigation domains and is designed as a reconfigurable primitive for agentic navigation systems, with consistent scaling from 2B to 8B parameters.

### xAI — Grok for PowerPoint

- Company: xAI (x.ai)
- Announced: 2026-06-16
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://x.ai/news/introducing-powerpoint-addin
- Record: https://forck.live/items/3452-grok-for-powerpoint
- Subject: Grok

Grok now works inside Microsoft PowerPoint, allowing users to create slides, expand decks, and refine narratives directly within PowerPoint.

### xAI — Grok Imagine Video 1.5

- Company: xAI (x.ai)
- Announced: 2026-06-16
- Category: model-update
- Coverage: not counted
- Announcement: yes
- Group: models
- Source: https://x.ai/news/grok-imagine-video-1-5
- Record: https://forck.live/items/3451-grok-imagine-video-1-5
- Subject: Grok

xAI announced Grok Imagine Video 1.5, an update that provides improved quality at even faster speeds.

### OpenAI — Predicting model behavior before release by simulating deployment

- Company: OpenAI (openai.com)
- Announced: 2026-06-16
- Category: research-paper
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://openai.com/index/deployment-simulation
- Record: https://forck.live/items/66-predicting-model-behavior-before-release-by-simulating-deployment
- Subject: GPT / ChatGPT / API

OpenAI introduces Deployment Simulation, a method to predict AI model behavior before deployment using real conversation data to improve safety and evaluation accuracy.

### Harvey — Building a Faster, More Reliable Upload Experience in Vault

- Company: Harvey (harvey.ai)
- Announced: 2026-06-15T15:00:00+00:00
- Category: capability-change
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://www.harvey.ai/blog/faster-more-reliable-vault-uploads
- Record: https://forck.live/items/17549-building-a-faster-more-reliable-upload-experience-in-vault
- Subject: Harvey

Harvey redesigned Vault's file upload system to use presigned URLs, moving the data plane from the backend to direct browser-to-storage transfers. The change addresses scaling bottlenecks that emerged as weekly upload volume grew from 2.2 million to 15 million files per week between January and May, and as Vault crossed 200 million active files. The new architecture splits control plane (authorization, validation, record-keeping) from data plane (byte transfer), reducing backend load and improving reliability for large batches.

### Swiss AI Initiative — Apertus Mini

- Company: Swiss AI Initiative (swiss-ai.org)
- Announced: 2026-06-15T09:00:00+00:00
- Category: open-weight-release
- Coverage: not counted
- Announcement: yes
- Group: models
- Source: https://apertus-ai.org/articles/2026-06-apertus-mini/
- Record: https://forck.live/items/18376-apertus-mini
- Subject: Apertus

The Swiss AI Initiative released the Apertus Mini collection, a set of 16 small models derived from the Apertus v1 large language model through distillation and quantization. The collection includes base and instruct versions at 0.5B, 1.5B, and 4B parameter sizes, along with multiple quantization levels. The models are optimized for memory- or compute-constrained systems such as portable or embedded devices.

### Factory AI — Factory 2.0: From coding agents to software factories

- Company: Factory AI (factory.com)
- Announced: 2026-06-15
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://factory.com/news/software-factory
- Record: https://forck.live/items/17835-factory-2-0-from-coding-agents-to-software-factories
- Subject: Droid

Factory announced the next phase of its platform, the software factory, an interconnected, agent-native system that triages signals from bug reports and feedback, then builds, tests, reviews, secures, ships, and monitors changes in a continuous feedback loop. The system is designed to be model-independent, sovereign, and capable of continual learning and self-improvement. Factory has been building software factories with customers including NVIDIA, EY, Adobe, Palo Alto Networks, Adyen, Blackstone, Wipro, and Comarch, and expanded visibility to manage the software factory directly in the Factory desktop app.

### xAI — Use Grok in Warp

- Company: xAI (x.ai)
- Announced: 2026-06-15
- Category: capability-change
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://x.ai/news/grok-warp
- Record: https://forck.live/items/3454-use-grok-in-warp
- Subject: Grok

xAI announces that Grok can now be used inside Warp, leveraging a Grok and X Premium subscription.

### xAI — Agent Dashboard in Grok Build

- Company: xAI (x.ai)
- Announced: 2026-06-15
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://x.ai/news/agent-dashboard
- Record: https://forck.live/items/3453-agent-dashboard-in-grok-build
- Subject: Grok

xAI announced a new Agent Dashboard feature within Grok Build that lets users manage multiple coding sessions, view their status, and send new work.

### Cohere — Cohere triples UK footprint with new London office to support R&D growth

- Company: Cohere (cohere.com)
- Announced: 2026-06-15
- Category: funding-company-news
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://cohere.com/blog/cohere-triples-uk-footprint-with-new-london-office-to-support-r-and-d-growth
- Record: https://forck.live/items/3278-cohere-triples-uk-footprint-with-new-london-office-to-support-r-d-growth
- Subject: Command

Cohere triples its UK footprint by opening a new London office to support R&D growth.