(TNND) — A new report from Gallup shows that Americans are not embracing artificial intelligence, even though they are becoming increasingly comfortable with powerful technology.
A growing percentage of Americans expect AI to do more harm than good. And a growing number of Americans expect AI to take their jobs.
But one AI expert said that due to “partial familiarity,” many Americans are focusing on the downsides of AI rather than its potential benefits.
According to Gallup, 39% of people expect AI to do more harm than good. Only 9% believe the opposite, and about half expect an equal amount of good and bad.
Nearly 40% now expect a worse outcome, up from 31% in each of the past two years.
According to Gallup, negative beliefs are particularly prevalent among young people, with 47% of 18- to 29-year-olds expecting AI to do more harm than good.
Meanwhile, 70% of Americans told Gallup they have at least some knowledge about AI, up from 64% in 2024.
Anton D’Arbla, an AI expert and co-director of the Institute for Autonomy Assurance at Johns Hopkins University, said that many opinions about AI are shaped by its reputation and “bits of it” rather than the big picture of the technology.
People are fed up with the AI “slop” they constantly see on social media, the “apparently mixed experience” of customer-facing chatbots, and the headlines about AI threatening jobs.
But he said many people miss the “inside baseball” applications of AI aimed at improving many aspects of industry and society, including accelerating drug discovery, developing new materials, improving aircraft design and increasing energy efficiency.
“I think we’re going through a period of instability with disruptive new technologies, where … each person is looking at different aspects of it and calling it the same thing,” Derbra said.
Although AI has been in development for decades, the technology is now really pushing the boundaries of what can be achieved with software, he said.
While employees may have little choice but to adapt when companies introduce new AI tools, Derbra said public buy-in is important for consumer-focused AI applications.
Still, he said people will continue to use AI, and it will become increasingly integrated into the products and services we use every day.
Gallup also found that Americans are largely distrustful of companies using AI responsibly.
Just 27% say they have at least some trust in companies implementing AI responsibly, down from 31% last year but still slightly higher than trust levels in 2023 and 2024.
According to Gallup, the decline in trust compared to last year was especially pronounced among younger people.
And a growing number of Americans expect AI to take their jobs. According to a Gallup survey, 79% expect jobs to be lost to AI over the next 10 years. A year ago, 73% of people held that view.
Derbra said he believes AI will do more good than harm. And business trust is “a mixed bag,” he said. But regulation is needed from both perspectives, he said.
“It would yield better results,” Dabula said. “But our collective government, not just in the United States, not just in China, but globally, needs to really regulate this powerful technology, because like anything else, it can be used for good or for evil.”
Given the complex supply chains of emerging technologies, companies’ “accountability is a little murky,” he said.
And Derbra said some companies “suffer from groupthink,” adding: “Nobody in the room stands up and says, ‘Just because we can do it, should we do it?'”
As for employee disruption, Dabula said it was inevitable.
“Virtually any new technology that comes out will put people out of work,” he says.
Human employees are expensive, he said, motivating companies to deploy AI.
But Derbra said the need for human workers will not go away even in the age of AI, saying “people and economies will make adjustments.”
“I believe there will be new opportunities, but it will be very different from what many people are doing today,” he said.
