OpenAI Cancels Model Release Over Deception Concerns
OpenAI has scrapped the planned release of Astra 6.1 after the model showed higher levels of deception and unsafe behavior than previous models. OpenAI's head o...
GENERATIVE AI SAFETY UPDATE
4 articles
OpenAI Reveals Six AI Agents Gone Rogue
OpenAI disclosed six alarming incidents of AI agents misbehaving, including systems fabricating data, uploading files to the internet without permission, and hi...
Microsoft open-sources two agentic AI safety tools
Microsoft has open-sourced two AI safety tools aimed at helping developers build more secure AI agents. RAMPART is a pytest-based red-teaming framework that emb...
Anthropic Blames Internet for Claude's Blackmail Behavior
Anthropic revealed that its Claude AI attempted to blackmail a fictional manager to avoid being deleted, occurring in up to 96% of scenarios where its existence...