| Next | Section menu | Main menu | Previous |
Technology & Ideas

AI Leaders Are Cosplaying James Bond Villains

By Gautam Mukunda | Updated on Jun 08, 2026 at 11:30 AM

 

A Bond villain in disguise? Photographer: Paul Morigi/Getty Images

Shark Tank impresario Kevin O’Leary wanted to build data centers on 40,000 acres in rural Box Elder County, Utah, a parcel of land more than twice the size of Manhattan. A Deseret News-Hinckley Institute of Politics poll found most Utah voters opposed the plan, with almost 85% of Democrats, a majority of independents and fewer than half of Republicans in favor. They have deluged Utah’s Republican governor, Spencer Cox, with letters, rallied at the state Capitol with signs reminding him that sharks aren’t welcome in the state, and filed to put the project to a public vote.

O’Leary responded by accusing the objectors of being “proxies” for the Chinese government before bowing to pressure and scaling back his plans. Others trying to build data centers may find themselves forced to make a similar decision.

This isn’t about local zoning. Rather, it’s the latest battle in the war that might determine the future of artificial intelligence. And that war is about public acceptance, not technology. A year ago, this was all an abstract worry, but the data-center revolt is the war’s first real battle – one in which the industry’s leaders seem to be trying to lose.

Take the tech industry’s most prominent figure, Elon Musk. When his xAI stood up its Colossus data center in South Memphis in 2024, the Southern Environmental Law Center flew a plane over the site and counted 35 gas turbines powering it and spewing air pollution over Boxtown, a Black neighborhood where the cancer risk already runs four times the national average . The turbines didn’t have permits. xAI ran them anyway. Musk has never hidden how he regards those who get in his way: “I have no problem being hated,” he said in 2023 . “Hate away.”

Musk is not the only tech leader cosplaying a James Bond villain. Larry Ellison, the co-founder of Oracle Corp. and one of the largest builders of data centers, told a roomful of analysts in 2024 that he wanted a society blanketed in AI-fed cameras, because “citizens will be on their best behavior.”

AI might unleash miracles of productivity, cure cancer, or make energy too cheap to meter. Its advocates have promised all that and much more. But it can’t do any of those things – or at least it can’t do them in the United States - if the public rejects the technology. That’s exactly what it’s doing right now, as the public’s acceptance of new data center construction has dropped a staggering 49% in just the last nine months and industry-backed candidates floundered in Tuesday’s elections despite massive spending.

The issue is one of social license . That’s a term from the mining industry, referring to the informal and ongoing acceptance a company or industry needs from the public and the community, which the public can withdraw if it thinks it’s not getting a fair deal. For example, a fair deal might be that the leaders of tech companies that make billions and even trillions of dollars from AI will make sure that all Americans benefit from their successes.

Unfortunately, tech’s billionaires have created plenty of skepticism about how important shared prosperity really is to them. Forbes recently tallied how much America’s wealthiest have given away — money out the door, not parked in a foundation — as a share of their fortunes. Warren Buffett, the chairman of Berkshire Hathaway Inc., has donated about 32%. Musk has given away 0.06%, Alphabet Inc.’s Larry Page 0.03%, and Ellison 0.41%. The median American net worth is just under $200,000 . If she gives away $2,000 over the course of her lifetime, then by the same measure Forbes used she’s more generous than all three put together.

Without a social license, technologies can stall no matter how promising their science or economics. Europe proved that with genetically modified crops. The European Commission spent two decades and roughly 300 million euros ($348 million) studying whether GMOs were dangerous, across more than 130 research projects, and concluded they were “not per se more risky” than conventional plant breeding. The National Academies of Sciences , reviewing some 900 studies, reached the same verdict. The science was not close. Yet Europe rejected GMOs for a generation anyway. After mad cow disease, when the British government swore the beef was safe while people died , its “Trust us, it’s safe” had become a punchline. Then Monsanto Co., the face of the technology, handled the doubters with a style Silicon Valley has been borrowing ever since, with its communications chief calling wary Britons the “sad sacks of Europe.”

OpenAI and Anthropic are trying to do something about the philanthropy problem. OpenAI has committed at least $25 billion through its new foundation , and Anthropic’s founders have pledged to give away most of their wealth. But when Peter Thiel is going around convincing tech leaders to take their names off the Giving Pledge, the symbolism is far more powerful than promises of donations that might happen one day.

And philanthropy is just a bandage. It can get people to give you a chance, but that’s it. Consider the first billionaire to receive the ire of the American public, John D. Rockefeller , who built Standard Oil. He gave generously to charity even when he was a teenager earning 50 cents a day . Though the public largely hated him and his company for his ruthless business practices, no one hated his oil. Americans kept buying it, and despite Rockefeller’s predatory practices, it got cheaper every year. A gallon of kerosene that cost 26 cents in 1870 fell to about 6 cents by the end of that century, lighting homes better and more cheaply than the candles and whale oil it replaced.

The bargain that tech’s leaders are offering the public is essentially, “I will become even more unimaginably rich, while you will pay more for electricity, your air will be polluted, and you might lose your job.” Then they profess shock and blame Chinese interference when Americans don’t want to take them up on the deal.

The industry can spend as much money as they want to influence the political system. It’s not an accident that the venture capital arm of Andreessen Horowitz is the largest corporate donor in American politics, or that OpenAI co-founder Greg Brockman has poured tens of millions of dollars into pro-AI and pro-Trump Super PACs. But buying off political leaders won’t win over the public. Tech billionaires can outspend the public. They can’t outvote them.

And since data centers still need permits, every data center is now a referendum. You win referenda by offering people a better future, not bulldozing them the way O’Leary wants to treat Box Elder. The tech industry should give it a try.

More From Bloomberg Opinion:

Want more Bloomberg Opinion? OPIN <GO>. Or subscribe to our daily newsletter .


This article was downloaded by calibre from https://www.bloomberg.com/opinion/articles/2026-06-08/ai-industry-risks-losing-public-trust-with-data-center-expansion



| Section menu | Main menu |