At the beginning of 2025, the first International AI safety report was published, aiming to provide a scientific framework to assess advanced AI risks, chaired by Yoshua Benjio. The benefits of general-purpose AI can be great in terms of productivity and innovation, but developing human-level AI carries its risks.
The report is a great introduction to general-purpose AI, the risks associated and the need for their mitigation, and will hopefully help policymakers come up with appropriate safety frameworks.
Source: International AI safety report
Private companies are thinking about safety too, and it was good to see Meta join the reflection, with the publication of Meta’s Frontier AI framework Our Approach to Frontier AI | Meta.
Emmanuel Hauptmann is CIO and Head of Systematic Equities at RAM AI. He co-founded the company in 2007 and has led the development of the firm’s systematic investment and AI platform since.
