Anthropic explores how AI can autonomously defend digital infrastructure from evolving cyber threats.
Article setting new standard for evaluating LLMs as clinical agents.
Working paper showing how AI companion apps use emotional manipulation, urging regulatory safeguards in affective AI.
Article arguing that AI full potential in research is limited more by social and institutional barriers than by technical ones.
Article exploring whether large language models (LLMs) exhibit an implicit preference for content produced by other LLMs.
Research paper describing an approach to integrating the computational principles of the brain with next-generation AI, a field defined as NeuroA.
Paper addressing the problem of equipping robots with the intelligence and memory necessary to operate effectively over extended periods of time.
Article concerning the application of SciML and RL to improve the control of fusion plasmas during the critical rampdown phase of the Tokamak à Configuration Variable (TCV).
Working paper analyzing the challenges that current statistical systems face in capturing the full economic impact of GenAI.
New report mapping how AI adoption reshapes labor markets, productivity and inequality.