On Wednesday, White House officials announced their work on a new executive order to oversee the release of frontier AI models. Today’s announcement follows on the heels of Anthropic’s Project Glasswing, a limited release of their latest AI model Mythos to prepare companies and the federal government for cybersecurity vulnerabilities. The new EO, previewed by National Economic Council Director Kevin Hassett, could create a vetting process requiring AI companies to receive approval from the government before releasing advanced models.
“The White House is driving at a key concept here: we need strong guardrails on advanced AI before it’s released, not just accountability when it breaks things,” said ARI President Brad Carson. “This is a promising step in the right direction and a fundamental shift in how the federal government is thinking about its role in de-risking advanced AI. For the White House’s oversight of AI to be effective, the Administration will need to set up basic measures including pre-deployment testing, mandatory standards, and audits. When our safety and national security is in Silicon Valley’s hands, voluntary commitments will not cut it.”
###
Americans for Responsible Innovation (ARI) is a nonprofit organization dedicated to policy advocacy in the public interest, focused on emerging technologies like artificial intelligence (AI). Learn more at ARI.us.
