Overview
Severity: MEDIUM | Affected: US AI Industry | Category: policy
The White House has signed the bipartisan "AI Security and Trust Act of 2026," marking a significant shift in US AI policy from voluntary industry commitments to legally binding regulation. The new law mandates that companies deploying "frontier" AI models in critical infrastructure sectors, including finance, healthcare, and energy, must undergo rigorous, independent third-party security audits before public release. These audits will assess vulnerabilities to adversarial attacks, data poisoning, and potential for misuse. The National AI Safety Institute (NAISI), under NIST, is tasked with certifying auditors and maintaining a public registry of audited models. Non-compliance carries substantial financial penalties, aiming to enforce a higher standard of safety and security for powerful AI systems integrated into society.