'Biological Risks May Soon Follow': Anthropic CEO Dario Amodei Warns AI Could Trigger Cyberattacks, Job Losses & Authoritarian Surveillance

Anthropic CEO Dario Amodei has published a sweeping policy essay warning that artificial intelligence (AI) is advancing far faster than governments can respond, and that the window to act is narrowing rapidly.

In his post titled 'Policy on the AI Exponential', Amodei compares the mismatch between AI's pace and policy speed to the Hobbits trying to rouse Treebeard in 'The Lord of the Rings', where one party moves at lightning speed and the other is fundamentally slow by nature.

The cyber threat is already here

Amodei points to Claude Mythos Preview, Anthropic's most advanced frontier model, as proof that the risks are no longer theoretical. He writes that frontier models now pose 'very real risks' to cybersecurity, with the potential for disruption of financial systems, critical infrastructure and national security.

"The cyber risks that Mythos-class models present will not be the last that we must face," he warned. "I believe that biological risks may soon follow, and that serious AI autonomy risks may not be far behind."

Jobs at risk

Amodei directly addresses the threat to employment. He acknowledges that despite best efforts, harm may be unavoidable. "Enduring job displacement is undesirable and dangerous. There's a decent possibility that, despite all our efforts, AI still causes significant enduring job loss."

He has called on governments to monitor AI's impact on employment and build support systems for affected workers.

AI as a tool of autocracy

Amodei also raises the threat of AI being weaponised by authoritarian regimes. "Powerful AI in the wrong hands could be the ultimate tool of autocracy. A surveillance-focused AI could analyse widely available information at massive scale and use it to infer the innermost details of every citizen's life," he said.

What Anthropic is calling for

Amodei argues the time for transparency-only policy has passed. Anthropic is now calling for binding regulation of frontier AI, mandatory safety testing, security requirements and government oversight before advanced systems are widely deployed. The company has also released a legislative proposal on frontier model testing alongside the essay.