FTC Probes AI Safety Failures at Leading Tech Firms The US Federal Trade Commission (FTC) has officially subpoenaed OpenAI and Anthropic to investigate recent AI safety failures. This regulatory action comes as the agency seeks to uncover potential consumer protection violations within the industry. Specifically, the commission is demanding documentation regarding security protocols, privacy policies, … Read more
Anthropic Grants METR Access to Curb Frontier AI Risks In a major shift toward transparency, Anthropic is granting employee-level access to six researchers from METR, a nonprofit focused on auditing Frontier AI models. This unprecedented arrangement allows independent experts to look under the hood of Anthropic’s systems to rigorously assess safety practices and risk management. … Read more
Anthropic Research Highlights Self-Replicating AI Risks Anthropic published a research report Monday warning that self-replicating AI systems could begin spreading across vulnerable Internet of Things devices within months. The frontier lab demonstrated that a 9-billion-parameter model successfully established inference operations on eight of nine hardware platforms tested, suggesting distributed deployment no longer requires centralized datacenter … Read more
