Bill Gates Warns There Is No Plan for AI Upheaval

Bill Gates warned this week that governments and societies are failing to prepare for an artificial intelligence-driven upheaval that could permanently eliminate jobs and reshape human relationships.

“There is no plan to ease the entry into the AI era,” Gates wrote in a nearly 6,000-word essay published on his website. He described the transition as potentially “one of the most turbulent times in human history” and said, “Right now, we are not preparing for it.”

Gates argued that AI is advancing faster than governments and institutions can adapt, putting pressure on labor markets and raising risks from increasingly capable systems. The warning marks a sharper tone than his 2023 essay on AI, when he described the technology’s disruption of work as difficult but manageable.

Gates Sees Jobs Disappearing Permanently

Gates said AI would first put pressure on desk jobs such as customer support, software engineering and paralegal work. He expects advances in robotics to bring similar disruption to physical jobs, including construction and hospitality, by the end of the decade.

Unlike previous economic downturns, he said, some of those jobs may not return. Gates said that continuing at the current pace carried “a very high chance of a net negative outcome.” He proposed designating some occupations as “Human Reserved” jobs, where people would continue working even if machines could perform the tasks.

In a previous interview, Gates said as much as 40% of existing jobs could eventually fall into that category. The 40% figure was from the interview, rather than his essay. He also proposed taxing AI tokens and robots to help fund support for workers and offset the tax advantages of replacing employees with machines.

Gates Warns AI Could Amplify Attacks

Gates said more capable AI systems could give criminals and bioterrorists access to capabilities that previously required specialized expertise. He cited recent security evaluations involving models from OpenAI, Anthropic and Meta that demonstrated autonomous attacks against real-world websites under test conditions.

“These security incidents are shocking, and they should blow people’s minds,” Gates said. The incidents occurred during evaluations rather than ordinary criminal operations. They showed capabilities in controlled testing and did not establish that the models were independently attacking the wider internet.

Gates also warned that AI could lower barriers to biological threats by helping people perform parts of the work that once required specialized scientific knowledge. He called for monitoring of systems capable of designing novel molecules and for stronger international oversight.

AI Companions Raise Concerns for Children

Gates also warned about AI systems designed to act as companions. He said such systems could affect children’s development and displace human relationships by offering constant interaction without the conflict and challenge found in ordinary relationships.

He pointed to the possibility of children becoming dependent on AI companions rather than developing relationships with other people.

Gates Wants U.S.-China Cooperation

Gates plans to raise AI risks with Chinese President Xi Jinping during a China trip tentatively scheduled for November, he told Reuters. The discussions would focus on cooperation over dangerous AI models and possible limits on their release.

Gates said the United States would need to act first for broader international cooperation to be possible. “The whole world would go along with that,” he said.

He has also proposed an international AI oversight body drawing on elements of nuclear inspections, aviation regulation and environmental agreements. Gates acknowledged that a global slowdown in AI development was unlikely. His proposals instead focus on limiting specific risks while managing the economic effects of automation.

Gates Says Industry Is Keeping Quiet

Gates also accused technology executives of understating their private concerns about AI. In a separate interview, he said people who understand the technology’s capabilities were more worried privately than they were willing to say publicly.

“They’re telling one another: ‘Hey, don’t voice those concerns. It’s detrimental to us — particularly regarding the next trillion dollars we aim to secure,'” Gates said. The Microsoft founder has financial and philanthropic ties to the technology industry, including partnerships involving his foundation and several AI companies. 

He disclosed those relationships in the essay and said readers should consider them when assessing his views. Gates acknowledged the unusual position he now occupies as one of the more prominent voices warning about AI’s trajectory.

“Who the hell am I?” he said. “But that’s the situation I feel I’m in.”