Third-party cyber evaluations involving OpenAI models

Simon Willison's blog post links to OpenAI's coverage of two third-party cyber evaluation incidents involving OpenAI models, one by the UK AI Safety Institute and another by Irregular, and notes Anthropic's write-up about Irregular's misconfigured environment affecting Claude.

From the source
This post from OpenAI covers both the UK AI Safety Institute attack (see my previous post) and another attack enabled by Irregular
simonwillison.net

Visit source ↗simonwillison.net

Announced: 05 Aug 2026 23:45 UTC
Detected: 06 Aug 2026 03:35 UTC (+3h after)

Other stories

xAI · 29 AugGrok Bot now works with XCapability change

GitHub · 28 AugGitHub Copilot in Visual Studio — August update

GitHub · 28 AugGitHub Copilot weekly releases — August 24Capability change

Amazon · 28 AugBatch write and discover records in Amazon SageMaker Feature StoreAPI release

Amazon · 28 AugHow Decathlon runs demand forecasting at scale with Chronos-2

Amazon · 28 AugSpreading the load: How Salesforce met Multi-AZ HA with SageMaker Inference ComponentsCapability change