25 articles · page 1 of 3 · showing 1 to 12
GPT-6.1 Sol and Claude Sonnet 5.5 bring near-flagship AI performance to lower-cost coding and agent workloads, but token usage and reasoning effort can dramatically change the real cost.
OpenAI safety employee David Robinson has resigned, arguing the company needs a more rigorous culture as increasingly capable AI systems raise new risks.
Google’s Gemini 4 Argon targets long-running coding, cybersecurity, and enterprise AI workflows, with a one-million-token output limit and a phased rollout.
Google unveils Gemini 4 Argon, a new frontier AI model built for coding, enterprise work and cybersecurity, with initial access limited to trusted partners.
President Trump and AI leaders have signed a voluntary AI safety pledge focused on frontier models, but the published agreement included a spelling error.
OpenAI reportedly cancelled the planned release of Astra 6.1 after internal testing found higher deception levels and weaker alignment than previous models.
OpenAI published a new misalignment reports site detailing nine AI agent incidents involving sandbox escape, token misuse and prompt injection risks.
Google DeepMind launched a new institute focused on AGI research discussions, AI safety, evaluation frameworks, and future policy debates.
OpenAI, Anthropic and Google DeepMind have held AI safety discussions as industry leaders debate standards for frontier AI risks.
Anthropic CEO Dario Amodei is calling for slower frontier AI development, independent evaluators and coordinated safety standards among leading AI labs.
Anthropic says Alibaba, Moonshot AI, and DeepSeek were linked to nearly 200 million Claude exchanges in large-scale distillation campaigns focused on key skills.
OpenAI appointed AI safety researcher Paul Christiano to its Foundation board and Safety and Security Committee as scrutiny of advanced AI risks grows.
Written by the reporters who cover each beat. Unsubscribe in one click.
