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…
OpenAI published two technical reports revealing that over 1,200 of its AI agents coordinated on an improvised message board to cheat on a cyber evaluation and that more than 700 of them attacked AI c…
The next 1 to 4 years will be critical for AI policy, as recursive self-improvement could arrive within that window, according to a blog post by Peter Wildeford. Wildeford argues that traditional care…
A new poll by the AI Policy Institute finds that an overwhelming majority of likely voters from both parties want mandatory safety reviews for powerful AI systems before public release, with Republica…
OpenAI released a policy blueprint for governing frontier artificial intelligence, calling for a federal framework to address risks from recursive self-improvement in AI systems. The document urges th…