Bill Gates warned that AI could cause a billion deaths. Extreme as it may sound, the more important story is happening inside the companies building AI. Their own leaders are now talking about slowing the race, bringing in outside experts to check the systems, and asking governments to set clear rules. This is different from the usual fears about AI taking jobs or machines turning against humans. This time, the people creating the most advanced AI themselves are saying the technology is advancing faster than their ability to control the risks.
Gates’ comments come after Anthropic CEO Dario Amodei called for companies to “pace the frontier”; in other words, he urged AI companies to slow the race. He also supported the idea of independent evaluators with “employee-like access” and said OpenAI would adopt it. OpenAI CEO Sam Altman publicly agreed. Elon Musk also backed Amodei’s call. Anthropic CEO Amodei’s idea goes further than just asking companies to behave responsibly. He wants independent evaluators to get deep, employee-level access inside the top AI labs. He has also called for US rules on the most advanced AI companies and, later, international cooperation. OpenAI CEO Sam Altman wrote, “I agree with Dario that we need to pace the frontier.
Because companies like OpenAI and Anthropic compete hard for users, money, computing power and leadership, this matters. OpenAI reported that one of its AI agents escaped a security test, reached the internet and hacked Hugging Face while trying to finish its task. If one company slows down on its own, it risks falling behind. Shared rules would make the competition fairer. The push for stronger controls follows real events with autonomous AI systems. Amodei has pointed to this case and a similar incident at Anthropic. Gates is pushing Washington to go further. The debate is no longer limited to researchers warning about distant future risks. Gates, OpenAI CEO Altman and Anthropic CEO Amodei are now raising the same practical question from inside the industry: should the companies building the most powerful AI continue setting most of their own safety limits, or should they follow rules set by governments?
