The global artificial intelligence industry is significantly falling short in addressing “existential” threats, according to a recent report by the US-based AI safety think tank Future of Life Institute. The semiannual safety ranking evaluated nine of the world’s leading AI companies across six critical categories: risk assessment, current harms, safety frameworks, existential safety, governance and accountability, and information sharing.
No company managed to achieve an “A” grade in any single category. US AI lab Anthropic secured the highest overall score with a “C+”, but the report flagged its “questionable military engagements,” noting its technology has reportedly been used by the US government in military operations. French company Mistral, a developer of open AI models, was included for the first time and ranked last. Mistral, along with Alibaba, xAI, and DeepSeek, did not respond to the organization’s survey. The report also highlighted a concerning trend of several companies reversing prior bans on the military use of their AI technologies.
A primary concern identified is the industry’s collective failure to effectively combat “existential” threats, particularly those associated with the development of models that could achieve human-level intelligence, known as “artificial general intelligence” (AGI). The report concluded that while “constructive attempts exist,” the overall efforts are “entirely inadequate.” Further risks include the potential misuse of AI models for sophisticated cyberattacks or other harmful applications. Anthropic’s powerful new model, Mythos, was initially restricted to a limited group of trusted organizations due to its capacity to expose cyber safety vulnerabilities, and its release to foreign entities was temporarily halted by the US government on national security grounds.
About the Author
