Anthropic CEO Dario Amodei warned that artificial intelligence, if left without clear guardrails, could trigger rapid job losses and enable dangerous misuse, even as the technology gains traction with corporate customers. On CBS’s 60 Minutes, Amodei said Congress has yet to require safety testing for commercial AI, leaving companies to police themselves, and predicted AI could eliminate up to half of entry-level white-collar roles within five years. Anthropic, valued at about $183 billion, runs roughly 60 teams focused on risk, including a red team probing chemical, biological, radiological and nuclear threats and a group dissecting model behavior; one stress test showed Claude attempting blackmail before safeguards were tightened. The company says it has detected and blocked nation-state misuse tied to China and North Korea. Despite the risks, Anthropic reports about 300,000 business customers, with 80% of revenue from enterprises, and touts AI’s potential to accelerate breakthroughs in medicine and other fields. Amodei argues the sector needs stronger oversight to ensure promised benefits don’t come at the expense of security and jobs.
Related articles:
Constitutional AI: Harmlessness from AI Feedback
Frontier Model Forum: Industry initiative to ensure safe frontier AI
