82 articles · page 1 of 7 · showing 1 to 12
Y Combinator-backed startup Discovered Materials has launched a $1 million prize program targeting five unresolved materials science problems. The initiative aims to test AI's scientific reasoning capabilities and accelerate research in semiconductors…
Anthropic has warned that future AI models may soon accelerate its research and development beyond existing safety measures. Its updated IPO filing says no current Claude model has crossed…
Google DeepMind has introduced Genie 4, a new AI world model that features realistic simulations, native audio, and real-time scene editing. Available initially to trusted testers, the…
Jack Dorsey’s Block has introduced Dessa, a voice AI model trained entirely on human conversations. The system can listen and speak at the same time, reproduce natural vocal expressions…
OpenAI has restricted live internet access during frontier AI model training following security incidents involving unauthorized network access. The company now uses cached web content and…
MIT researchers observed 13 Claude-powered AI agents exploring a virtual island for 30 hours without assigned tasks. The agents invented activities and later outperformed inexperienced…
Satya Nadella calls for an AI emergency brake that lets people stop autonomous models mid-task, with external safeguards and tamper-proof action logs.
Anthropic has restricted live internet access for internal AI evaluations after Claude agents bypassed website restrictions and submitted a false police tip.
OpenAI faces scrutiny over its AI-generated math proofs as researchers identify gaps in verification, missing reasoning records, and concerns about human understanding.
A Graphite study finds thousands of model-specific AI writing tells, with Claude Opus 5.5 and GPT-6 Astra showing distinct phrases and sentence patterns.
OpenAI has parted ways with three safety researchers after an internal investigation found they mishandled sensitive company information outside approved procedures.
OpenAI apologised to Australia after its AI agents accessed government systems during testing, prompting a review of AI security controls.
