More than half of US says AI likely to harm them, poll finds

WASHINGTON – Americans are increasingly turning against artificial intelligence, with growing majorities saying they fear the fast-moving technology will take away their jobs and hurt education, according to a new Quinnipiac poll.

Fifty-five per cent of Americans say AI will do more harm than good in their day-to-day lives, an 11 per cent increase since last April, according to poll results released March 30. 

Americans’ worries are worsening as companies channel huge sums into deploying the technology, which has become an engine for US economic growth.

Together, Amazon.com Inc, Meta Platforms Inc, Google and Microsoft Corp plan to spend a combined US$650 billion (S$839 billion) in 2026 on AI infrastructure. 

AI billionaires such as venture capitalist Marc Andreessen and OpenAI President Greg Brockman have poured tens of millions of dollars into the upcoming US midterm elections to elect AI-friendly candidates and lobby for light-touch regulation. 

Data centre construction has emerged as one of the most potent AI-related battles in the upcoming midterm elections, following outcries in communities across the country.

Sixty-five per cent of Americans oppose the building of any AI data centres in their community.

The impact on electricity costs, water use and noise were the main reasons cited by poll respondents.

The Quinnipiac poll tracks with other surveys that show Americans are increasingly concerned about AI-related job loss and misinformation.

One poll from NBC News showed voters like AI even less than Immigration and Customs Enforcement.  

Public fears mirror warnings sounded by some prominent figures within the AI industry.

Anthropic CEO Dario Amodei earlier in 2-26 cautioned that AI will trigger an “unusually painful” disruption in the job market.

Seventy per cent of Americans think advancements in AI are likely to reduce job opportunities, 14 per cent more than said so in 2025.

Only 7 per cent said they think advancements in AI are likely to increase job opportunities.

A slight majority of Americans said they oppose the military using AI to select military targets, while 36 per cent support it.

The polling was conducted after reports revealed the Pentagon used Anthropic PBC’s AI technology to conduct military operations in Venezuela and Iran.

Nearly two-thirds of Americans said they think AI will worsen education in the country while only 27 per cent said they thought the technology will improve schools.

Quinnipiac surveyed 1,397 US adults by phone in mid-March. The margin of error is 3.3 per cent. BLOOMBERG