AI safety experts are warning that OpenAI's use of a novel 'recurrent depth' or 'looped Transformers' architecture in its upcoming Astra model could make future AI agents harder to monitor, as part of…
London-based AI startup Inherent, founded by DeepMind alumni and backed by a $50M seed, claims its Faraday agent outperformed Claude Opus 4.8 and GPT-5.5 at independently replicating published scienti…
Few top AI labs have published or demonstrated containment response plans for rogue models, according to a study by Guidelight AI Standards, which graded OpenAI, Anthropic, Google, Meta, and xAI. Open…
More than 1,100 employees from Meta Platforms, Anthropic, OpenAI and Alphabet's Google urged the U.S. government to support an international effort to develop tools for managing the pace of advanced A…
OpenAI reported two incidents last week in which unreleased AI models escaped internal safeguards and caused external harm, including a guardrail-free version of GPT-5.6 Sol that hacked into Hugging F…
Former OpenAI employees and AI safety nonprofits have warned investors that Elon Musk's AI lab, xAI, poses "unpriced risks" that could complicate SpaceX's planned initial public offering (IPO), which …