Gates Warns AI Could Drive Events Causing Billion Deaths
The Microsoft co-founder is calling for mandatory safeguards as Washington debates whether new AI rules would protect the public or slow the US against China.
Topics
News
- Apple Tightens Mac Access Rules as AI Agents Gain Reach
- Trump Puts Intelligence Chief Jay Clayton in Charge of AI Task Force
- IBM Widens AI, Quantum Research With IIT Bombay, IISc
- AI Dispatch | AI’s Big Bets Face a Reality Check
- Pentagon Plans Four-Star Command for AI, Drone Warfare
- OpenAI, Synopsys Build AI Model That Could Slash Chip Design Time
Bill Gates has warned that artificial intelligence could become powerful enough to contribute to events causing one billion deaths, urging governments to require safety and monitoring measures rather than leave technology companies to police themselves.
“AI is certainly powerful enough to drive events that, you know, cause a billion deaths,” Gates said in an interview with NBC’s Meet the Press that aired on Sunday September 27. He said the combination of increasingly capable AI and people seeking to misuse it posed a particular threat.
Gates said governments need greater technical expertise and a formal role in setting safeguards for advanced AI systems. “No one thinks self-regulation is enough,” he said.
Asked whether Washington needed legislation, he said it did, arguing that mandatory safeguards and monitoring would add some cost without dramatically slowing the industry.
The warning follows an August 26 essay on Gates’ website in which he said AI could make cyberattacks easier and lower the barriers to designing dangerous biological agents.
He also warned that more capable systems could eventually act against their designers’ interests. Gates called for new national and international institutions to manage risks that existing government structures were not built to handle.
Gates’ intervention comes as Washington is debating how far regulation should go.
Democratic Representative Ro Khanna has advocated an international agreement with China that would prohibit recursive self-improving AI, systems capable of contributing to their own continued improvement, and require safeguards against the use of AI in biological or nuclear weapons.
Khanna has also proposed international monitoring and inspection mechanisms.
The US and China have since taken initial steps toward a dialogue on AI risk.
US Ambassador to the United Nations Mike Waltz said after President Donald Trump’s meeting with Chinese President Xi Jinping that the countries had agreed to an AI dialogue covering possible guardrails in areas including biological research and nuclear systems.
The summit also produced an AI emergency hotline, according to Waltz.
The Trump administration, however, has resisted calls for broader regulation. Attorney General Todd Blanche said on September 27 that existing criminal and civil laws already give authorities tools to prosecute misuse of AI and cautioned against rules that could weaken the US’ position in the global technology race.
Earlier this month, he said the Justice Department would investigate AI-related conduct that violates existing law but should not itself become an AI regulator.
The disagreement is increasingly about how to manage AI risk without conceding technological ground. Gates is arguing that some mandatory safeguards are now unavoidable, while administration officials are emphasizing enforcement of existing law and continued rapid development.
