Gates Pushes China, US for AI Bioweapon Safeguards

Bill Gates is seeking a meeting with Chinese President Xi Jinping later this year to press for coordinated global rules on artificial intelligence, including monitoring systems to prevent the technology from enabling biological attacks. The Microsoft co-founder described the effort as a response to risks that differ sharply from those posed by earlier technologies, and he said he expects China to back restrictions on the release of hazardous AI models if the United States moves first.

In remarks to Reuters and in a lengthy new essay, Gates outlined a shift in his assessment of AI. He had previously emphasized the technology’s potential benefits, but he now argues that its speed and scope require new forms of policy oversight that existing institutions are not prepared to provide. Gates said he intends to tell Xi that nations should track AI’s capacity to synthesize molecules and carry out biological attacks, measures he believes would not slow legitimate development of the sector.

Gates last met Xi three years ago, becoming the first foreign entrepreneur hosted by the Chinese leader after the pandemic. His staff is now working to arrange another session during a planned November visit to China. He acknowledged that Xi might be surprised by the strength of his concerns, noting that while China has taken steps to limit children’s exposure to AI companions, few governments have addressed the wider range of dangers.

The proposal for joint monitoring comes as Gates warns that AI could become either the greatest force for equality or a major source of new inequalities, depending on how societies manage the transition. He has described the current period as more disruptive than past technological shifts, such as the spread of personal computers, and said misleading comparisons to earlier innovations have left many policymakers underestimating the pace of change. Gates added that he would likely support any credible plan to slow AI advances worldwide if one were available.

His recent essay expands on these points and calls for governments to confront the possibility of cyberattacks, large-scale job displacement, and growing psychological dependence on AI systems. Gates has said the world is not preparing adequately for these outcomes, even as companies continue to release more capable models. The essay marks a departure from his earlier public stance and reflects a broader pattern among some technology leaders who once downplayed regulatory needs.

The effort to engage China directly on these issues highlights the difficulty of building effective international standards. Any agreement would require alignment between the two largest AI developers, yet tensions over export controls, data access, and model safety already complicate cooperation. Gates has expressed optimism that shared interest in preventing catastrophic misuse could provide a starting point, but he has also noted that most global leaders have yet to treat AI risks with the urgency he believes they warrant.

If the meeting occurs, it would represent one of the more concrete attempts by a prominent technology figure to shape cross-border policy at a moment when national approaches remain fragmented. Gates has indicated he plans to focus the discussion on practical safeguards rather than broad calls for restraint, arguing that targeted monitoring can address the most severe threats without halting progress.