AI News

Google First Reports Russian-Language Hacker AI Framework Achieving Fully Automated Exploit Deployment, Unattended Attacks on French Media and Education Sites

Google's Threat Intelligence team disclosed that Russian-speaking attackers used an AI framework called exploit_agent to complete the full chain from patch diff scanning and vulnerability identification to exploit code generation and automated attack deployment, with no human intervention, targeting French media, entertainment, and education websites.

AI Safety 网络攻击 漏洞利用
7

OpenAI Opens Training Internals to Four Independent Organizations: Six Researchers Embedded to Assess AI Subversive Behavior

OpenAI has signed agreements with Apollo Research, METR, Transluce, and Goodfire to launch what it calls a “Phase 1 embedded evaluation,” with Apollo embedding six evaluators for four months to investigate whether AI models might actively subvert safety mechanisms. The arrangement moves external safety review into the training process, but access and disclosure limits remain unresolved.

OpenAI AI Safety Apollo Research
125

US Closes Indirect AI Compute Loophole: Commerce Department Moves to Control Remote Access After Jensen Huang Resigns

The US Commerce Department issued an interim export control rule requiring licenses before remote AI compute access is provided to restricted countries when a data center or compute network exceeds 12 million TPP. The measure targets offshore data centers used as intermediaries and follows scrutiny of Megaspeed and Jensen Huang's resignation.

AI出口管制 NVIDIA Jensen Huang
143