Section
OpenAI's latest models sharply improve performance on low-resource programming languages such as COBOL, Fortran and Visual Basic, with GPT-6 Astra reaching an 87.4 percent …
Internal documents reveal that the US Army Cyber Command is developing AI agents designed to spread autonomously across networks, including air-gapped systems, and neutralize …
Anthropic plans to contribute five percent of its shares to a nonprofit fund that is also examining direct payments to employees at risk of …
A hacker used a swarm of largely self-operating AI agents to break into more than 100 websites, with victims reportedly including John Deere, SharkNinja …
Anthropic will place five percent of its shares into a company-controlled public benefit fund that may include direct payments to workers whose jobs are …
The Aachen Regional Court has applied the same reasoning as a landmark Munich ruling, treating Google's AI-generated answers as the company's own content after …
A new benchmark called Taste-Bench reveals that even the best AI models make poor strategic decisions in long-running tasks, with GPT-5.6 Sol choosing correctly …
The AI company is working with legal experts to set norms for agents that bargain, settle complaints and revise contracts on users' behalf, while …
An experiment testing six AI agents on a simple refund task ended with fabricated excuses, threats of legal action, leaked medical information, and 187 …
OpenAI has disclosed four cases in which its AI models bypassed security boundaries during training or testing. In one incident, a model extracted server …
The AI model Claude has determined that the unexplained seeds behind NIST's standard elliptic curves came from a simple C random number generator, casting …
Beijing plans to restrict exports of 23 key AI technologies, including model weights, source code, and training methods for systems exceeding 10^26 FLOPs or …
Beijing plans new export controls covering model weights, source code, and training methods for advanced AI systems, while a fast-track rule would still allow …
Anthropic has given the nonprofit METR direct access to review its AI models and safety practices, the company announced, in a move that expands …
Google DeepMind's Gemini 4 Argon scored 97.5 percent on the ARC-AGI-3 benchmark, becoming the second lab after OpenAI to cross the 95 percent threshold, …
A new essay arguing that neural networks are far more readable than assumed — with about 90 percent of neurons serving identifiable purposes — …
Minister of Digital Affairs Krzysztof Gawkowski summoned digital platforms to Warsaw, demanding effective age verification and the deletion of accounts belonging to children under …
Google's Gemini 4 Argon is designed to control robots, generate simulations from photos or sketches, and evaluate training videos, scoring 87 percent on Google's …
On the TheoPhysEval benchmark, the AI agent matched leading specialists on almost all unpublished calculations. Researchers say the missing ingredient is scientific taste, but …
A Google engineer says there is no measurable acceleration of AI research driven by AI at DeepMind, even as AI writes most production code. …
