Anthropic and OpenAI Push for AI Safety Rules

Anthropic and OpenAI CEOs are publicly warning that advanced AI systems pose dangers and require regulation, while critics question whether their motives include political influence and financial gain ahead of stock listings. The companies are framing safety debates to position themselves as cautious leaders in an unregulated market. The executives have called for independent testing and oversight, citing scenarios of uncontrolled AI, but experts note they avoid addressing immediate harms like military use or surveillance. The Trump administration has dismissed existential risk warnings, while a federal agency exists for voluntary model evaluations without universal safety standards. The companies are promoting their own auditing parameters and selected evaluators, potentially creating barriers for smaller competitors and appealing to investors. Critics argue this redirects political will away from meaningful regulation, while some former employees continue urging pauses on development to prevent catastrophic outcomes.