A majority of U.S. voters do not want the government using artificial intelligence (AI) to make decisions, according to a new poll from Americans for Responsible Innovation (ARI) shared first with The Hill.
Fifty-five percent of registered voters said they disagree with the statement that the government should use AI to make decisions about eligibility for unemployment assistance, college tuition aid, research investments, food aid and small business loans.
Another 31 percent agreed with the government using AI in these decision-making processes, while 15 percent said they didn’t know or had no opinion.
The fears surrounding AI may partly be a result of a lack of understanding about the technology, said Eric Gastfriend, executive director and co-founder of ARI, an AI policy advocacy group.
Nearly a fifth of voters in the survey said they had never heard of ChatGPT, one of the most popular AI chatbots.
“I think people are concerned about it partially because they don’t know what it is,” Gastfriend told The Hill. However, he added, “I think also people have legitimate fears and concerns about how the technology can be misused in ways that are discriminatory.”
These fears tend to manifest differently across the political spectrum, Gastfriend noted. On the left, these concerns center around potential discrimination based on someone’s identity, while the right is more concerned with discrimination based on viewpoint.
“There may be able to be some bipartisan room for policymakers to work together on ways to create some safeguards and protections around this technology,” he said.
Just under half of voters in the poll — 46 percent — agreed with the statement that the government “should slow its integration of artificial intelligence to avoid potential harms to taxpayers, employees, and the work of federal agencies.”
Nearly a third said the government was adopting AI at an appropriate pace, while 22 percent said it should act “more quickly to integrate” the technology.
“I think first we need to recognize the limitations of AI,” Gastfriend added. “At ARI, we’re big believers in AI, and it’s an incredibly powerful and transformational technology. But there’s a lot of stuff that I can’t do and sometimes the hype gets ahead of that.”
The concerns about government AI use come as reports have emerged that Elon Musk’s Department of Government Efficiency (DOGE) is considering potential uses for the technology amid its push to slash government spending.
Almost six in 10 voters in the survey also said they are concerned the government will not go far enough in regulating AI chatbots as they proliferate, and nearly eight in 10 said they are concerned about the impact the technology will have on younger generations.
The ARI poll was conducted by Morning Consult between Feb. 13 and Feb. 15 with 2,076 registered voters and has a margin of error of 2 percentage points.
