Marc Andreessen Defends Targeted AI Regulation as US Restricts Anthropic Model Access

Marc Andreessen published an article on June 15 defending a narrower approach to AI regulation focused on safety and trust rather than broad bureaucratic restrictions. The A16z co-founder argued that overly restrictive frameworks risk stifling innovation, entrenching incumbents, and slowing technological deployment, while functional regulation—such as safety standards, fraud prevention, and market safeguards—serves as enabling infrastructure.

The comments coincided with escalating US controls on frontier AI systems. The Trump administration has restricted foreign access to advanced Anthropic models through export controls, preventing foreign governments, companies, and individuals from using them. US Commerce Secretary Howard Lutnick informed Anthropic CEO Dario Amodei that specific models would fall under export restrictions following concerns about potential security vulnerabilities. The restrictions remain in place pending additional safeguards.