Technology
Archived — This article has been archived. The information may be outdated.
Anthropic CEO backs government AI model blocks
NDTV Profit··11 Jun
Anthropic CEO Dario Amodei said governments should be able to block AI developers from deploying new models if they present unacceptable risks. He called for mandatory third-party testing across categories including cybersecurity threats and biological weapons risks.
Prism
What It Means For You
- Frontier AI regulation may shift from voluntary transparency to formal testing and deployment approval.
- Government veto power could slow risky models, but it also raises questions about innovation, capture and political misuse.
- Third-party testing can help users trust AI systems if test standards are public, rigorous and independent.
What's Happening
- Dario Amodei said governments should be able to block dangerous AI model deployments.
- He called for mandatory third-party testing for risks such as cyber and biological misuse.
- His comments contrast with a more hands-off US executive approach focused on voluntary model access.
Frontier AI Governance
- Advanced AI models can improve productivity but may also assist cyberattacks, misinformation or dangerous scientific misuse.
- Regulators are trying to decide whether AI should be governed like software, infrastructure, medicine or aviation.
- Companies that build frontier systems increasingly face pressure to prove safety before broad release.
all-newstop-stories
