All news
In the news · CNN ·
Anthropic CEO tells CNN how AI ‘agent swarms’ could threaten humanity
Anthropic’s CEO warns CNN that autonomous agent swarms could threaten humanity. Frontier systems this powerful need human accountability—not slogans.
When the CEO of a frontier lab tells CNN that agent swarms could threaten humanity, that is not a reason to freeze American research. It is a reason to treat autonomous systems as a national-security problem.
Swarms of agents that can act without a human in the loop fail the test we apply to legislation: meaningful human accountability for consequential decisions. Panic is not a policy. Neither is “move fast and hope.”
American AI Safeguards will score the record on evaluation, biosecurity, cybersecurity, and human control—in public.