Technology is now more unpopular than ICE, the immigration police. A reality that transcends political divisions and leads AI bosses to change their anxiety-provoking discourse.
On May 15, Eric Schmidt gave a speech at a graduation ceremony at the University of Arizona. While he told students that the technological transformation brought about by AI will be “bigger, faster and more consequential than anything that has preceded it,” the former Google boss was greeted with a chorus of boos from the students.
If the event may seem anecdotal, it is in reality representative of the increasingly negative vision that the American public generally has. instead technophileAI. Polls show it one after the other: Americans are extremely skeptical of this technology.
A transpartisan and transgenerational rejection of AI
A study by the Pew Research Center, carried out in September 2025, shows that 50% of adults are more worried than enthusiastic about AI. Only 10% think the opposite. According to a poll conducted by NBC News in March 2026, 57% of American voters say that the risks posed by AI outweigh its benefits. These results are all the more striking since most of the AI giants are American: it therefore does not pose a risk of dispossession, loss of sovereignty and technological vassalization for the American public, as is the case in Europe.
Even more worrying for the AI barons: this opposition clearly transcends political divisions. At a time when Silicon Valley gets closer of Donald Trump, one might expect opposition to AI to emerge primarily from Democratic ranks, while otherwise pro-business Republicans are supportive of the technology. This is not the case. A poll carried out by the University of Maryland in August 2025 shows powerful bipartisan support for the regulation of AI, an area in which the Trump administration is currently conspicuous by its absence: 84% of Republicans and 81% of Democrats want the law to require companies to have their AI tested by the government before being able to deploy them.
Finally, this skepticism particularly affects the youngest. A poll co-conducted in April by Gallup, the Walton Family Foundation and GSV Ventures shows that only 22% of Generation Z Americans are enthusiastic about AI, compared to 36% a year ago. A result to be related to the difficulties that young Americans are currently experiencing in the job market, difficulties that many attribute to the adoption of AI by companies, which would replace junior positions, although some economists dispute this hypothesis.
Growing opposition to the establishment of data centers
This rejection of AI by the Americans, and the headache that it poses for the CEOs of Silicon Valley, is embodied in particular in a frank opposition to projects for the establishment of data centers, in which the AI giants are investing with all their might to acquire more computing power and feed their algorithms. 71% of Americans are opposed to the construction of one near their home, more than those who oppose the construction of a nuclear power plant nearby.
Voters have also begun to walk their talk and mobilize across the country. HAS Festus, Missourivoters ousted four City Council members a week after it approved the construction of a $6 billion data center. Galvanized by this example, dozens of municipalities, spread across numerous states, are seeking to prohibit the construction of new data centers, a real danger for big tech in a country where local democracy is much more powerful than in France.
In total, collective mobilization blocked or delayed about 50 projects worth about $156 billion last year, according to Data Center Watch, an organization that tracks local opposition to data centers. A record 20 projects were canceled in the first quarter of this year alone due to opposition from local populations, according to figures from climate media outlet Heatmap. Dozens of others are currently facing similar obstacles.
Politicians are also starting to take up the subject. Republican Senator from Missouri, Josh Hawley, known for his populist line and his opposition to Big Tech, proposed, with Democratic Senator from Connecticut Richard Blumenthal, a law aimed at requiring data centers to find independent energy sources to avoid weighing on the electricity grid and Americans’ electricity bills. South Carolina Republican Rep. Nancy Mace has proposed a one-year moratorium on the construction of new data centers in her state, while a Texas lawmaker, also a Republican, wants to ban any new development of AI data centers in the state, citing concerns about costs for farmers and pressure on the power grid.
The revolt sometimes also takes more violent forms: in Indiana, shots were fired at the front door of a municipal councilor who had recently approved the construction of a data center. Sam Altman’s home was targeted with a molotov cocktail.
A PR problem
Beyond the concrete obstacles to the installation of infrastructure necessary for the development of AI, the fact that the technology is highly unpopular is a big problem for Silicon Valley. Several experts have long been sounding the alarm on this subject, pointing to the anxiety-provoking speech of AI bosses and its responsibility for the growing public hostility towards it. “‘Our product will make you economically useless and maybe even kill you’ is not a very selling point,” wrote American economist Noah Smith last March.
According to him, the AI barons have overplayed the transformative aspect of their technology to attract funds from VCs, and now find themselves trapped by a pitch that does not sell well in the eyes of the public. “In the early days, when AI desperately needed capital to scale up, wooing companies and investors with the promise of lower labor costs may have been a good communications strategy, especially since most of the country wasn’t paying attention.
But today that is no longer the case, and founders of AI companies need to understand that the general public in America, and probably most other countries, thinks about economic well-being first and foremost in terms of jobs. Very few people are ready to become life-long welfare recipients, with all that that implies in terms of dependence, precariousness and lack of prospects.”
Donald Trump himself recently acknowledged that AI needs better PR. Aware of the problem, AI bosses have started to change their tune. “We want to build tools to augment and elevate people, not to replace them,” Sam Altman posted on May 1 on his Jensen Huang recently strongly criticized his colleagues linking AI to job destruction.
The fact remains that the risks that these entrepreneurs have highlighted for years (the existential risk posed by AI, the threat of mass unemployment, autonomous weapons, etc.) are not going to disappear overnight in the minds of the public, and that it will take more than an adjustment of the commercial pitch to convince them that AI will bring more good than harm.
