red teaming - Page 4
Stories from red teaming come first, newest first, then stories from connected topics (alignment, AI Safety, Codex, AI evaluation, AI Policy & Safety, Claude), each labelled with where it comes from. This view is not indexed by search engines.
Latest
OpenAI Bug Bounty Payouts Jump as Cathedral Joins Its Security Sprint
OpenAI has more than quadrupled typical bug bounty rewards, with up to $172,000 for exceptional finds, and brought in Cathedral for a security sprint.
Anthropic’s FlashProve Uses Cryptography to Prove Which Model Replied
Anthropic's open-source FlashProve prototype proves a reply came from a committed AI model without revealing its weights, at about 70-95x the GPU cost.
CrowdStrike Pairs AI Kill Switch With Deals for Fiddler and Dreadnode
The CrowdStrike AI Kill Switch lets admins stop rogue AI agents. CrowdStrike also agreed to buy Fiddler AI and Dreadnode; prices were not disclosed.
OpenAI Says Activation Probes Now Watch Its Models as They Answer
OpenAI says activation classifiers now scan its top models' answers token by token. It reports better jailbreak resistance but stresses clear limits.
Gemini May Breach Press Publishers’ Rights, Says EU Court Adviser
An EU court adviser says Gemini and press publishers' rights can clash when AI answers reuse paywalled or recognisable articles. It is not a ruling.
Trail of Bits Says Codex Helped Get 149 CVE Fixes Into Stable Linux
Trail of Bits says backporting security fixes with Codex cut the work from days to hours, helping land fixes for 149 CVEs in stable Linux kernels.
UK Expands DeepMind’s PlanAssist Planning AI to 15 Councils
PlanAssist, the AI planning tool built with Google DeepMind, expands to 15 English councils. The government says it halves routine householder decisions.
ChatGPT and Codex Will Share One Usage Allowance From December
From December, ChatGPT and Codex will draw on one token-based usage allowance for Plus, Pro and Business users, OpenAI says. Instant chat won't count.
Grok API Text Will Carry an Invisible Watermark From 7 November
The Grok API watermark arrives on 7 November: text from seven Grok models gets an invisible mark by default, citing the EU AI Act. Admins can opt out.
Anthropic Steers Claude Code Away From Cloning Existing Games
Anthropic has added a Claude Code rule on Opus 5.5 against recreating specific existing games. Original and genre games should keep working, it says.
Latest on TSN
Explore the TSN Atlas: Every topic we cover, mapped. Click a bubble to dive in.
