Can AI gain widespread acceptance if vast swaths of the public don’t trust it? A new poll by the Edelman Trust Institute suggests the public has trust issues when it comes to AI, a finding that may have ramifications in the US in particular as the country heads into the mid-term elections in 2026.
A key finding is that high income earners feel themselves at risk from AI and more than half would support an AI safety net for people who lost their jobs to AI. Another key finding that may be counter to expectations is that only 4 in10 young people between 18 and 34 years old trust AI.
The distrust of AI also seems to run along a global divide. Less than 18% of people in the U.S., Great Britain, and Germany said they fully embrace AI. By contrast, AI is embraced in China by 54% of the population, followed by Brazil at 35%.
AI acceptance also varies by industry. As might be expected, AI acceptance in technology and finance sectors is strong. However, AI acceptance lags significantly in many sectors like retail, manufacturing, healthcare, education, food & beverage and transportation with levels below 35%.
So what is driving distrust of AI? In the U.S., 70% cited data protection as their major concern. Developments like the rise of dynamic pricing based on AI algorithms that analyze consumer financial information to set prices is bound to make data protection a high profile issue going forward as the traditional notion of “one price for all” is undermined. In New York, a newly enacted law policing algorithmic pricing that “reads people’s minds” is the first in the United States. The New York initiative is cited as a model for similar AI guardrail legislation in other states. The absence of federal guidelines is turning AI into a local issue that may determine political careers. Adding to the distrust is the oft-cited tendency of AI to dispense hallucinatory misinformation like an oracle from ancient Greece gone mad.
Job anxiety also drives AI distrust. In both the U.S. and UK, 70% felt that employers were not being honest about job cuts caused by AI. That anxiety about employment may be well-founded. Last week, the Massachusetts Institute of Technology (MIT) released a study that found AI is already in position to replace 11.7% of the labor force to the tune of $1.2 trillion in wages. MIT’s new “Iceberg Index” is able to analyze AI labor disruption down to individual zip codes. The Edelman survey suggests that long term job security dramatically increases the likelihood of workers embracing AI. More than half of U.S. and UK respondents fear being left behind by AI.
Worries about AI seem to be coming to a slow boil. A Pew Research study in September revealed 50% of Americans were more concerned than excited about AI use in daily life, up from 37% in 2021. The erosion of creative thinking and an inability to spot misinformation like AI-generated “slop” on social media were among the chief concerns.
A caveat with political implications is that roughly 60% of the Edelman participants support an income safety net for those who lose their jobs due to AI. In the same vein, only one-quarter view U.S government leaders and regulators as trusted voices regarding AI. Some politicians already are seizing upon this deficiency as an opportunity. Increasing resistance to skyrocketing electricity rates associated with the proliferation of AI data centers is upending the political calculus. For example, two Democrats last month were elected (in a landslide) for the first time since 2007 to the Georgia Public Service Commission, the “red” state’s energy regulator. AI data centers planned by Amazon in Tucson, AZ and Google in Indianapolis, IN have been shelved due to local resistance. In Virginia, home to the most AI data centers in the U.S at over 600, electricity costs were a central issue to the election of incoming Democratic Governor Abigail Spanberger who vowed AI companies will “pay their fair share.”
Similarly, in Virginia, Democrat John McAuliff flipped a seat in the state legislature with a door-to-door “affordability” spiel focused on data centers and rising electricity costs that “were essentially an artificial tax on everyday Virginians to benefit Amazon, Google.” McAuliff also told The Guardian “which is not to say they don’t provide benefits to those communities but we need to do a much better job of extracting those benefits because companies can afford them.”
While AI distrust looms high, people also are optimistic in some regards. A majority expect to trust AI with help in making purchasing decisions, health management and job searches. Seven in 10 believe AI will help solve societal issues related to climate change. Of course, that too is a political issue, especially in the eyes of the Trump Administration which views climate change as a hoax.
See you on the campaign trail in 2026, AI!
