Artificial Intelligence: Key insights, data and tables

The large language model (LLM) revolution has brought an explosion of innovation in artificial intelligence (AI) and related technologies — the social and cultural impact of which remain to be seen. Ipsos is tracking the issues around data, trust, and adoption, in the U.S. and worldwide. Read on for a high-level look at the latest data — or visit "Gen AI: From Wow to How," the latest installment in our Insights to Activate series.

Key takeaways:

People are worried about biased AI — but not as much as they're worried about biased people
- When it comes to implementations of AI, algorithmic bias is a frequently cited concern. Ipsos' data suggests people are worried about the misuse of AI in various sectors — but that we trust AI not to be biased slightly more than we trust people not to be biased.
Interest in AI solutions varies by category
- When AI-powered image generators hit the market, millions were astonished by their creative potential. ("AI, generate me a picture of a penguin in the style of Degas.") But Ipsos trendlines indicate that people are getting to be less enthused about the use of AI in art, TV, and media — and that's to say nothing about how trust varies when it comes to healthcare and other sectors.
We have mixed feelings about AI in the workplace
- We know that AI is already being put to use at work, with nearly half of workers (46%) in the United States, United Kingdom, and Australia saying they've used these tools on the clock. But that hasn't assuaged fears about job security — particularly in labor-intensive sectors like manufacturing.
Americans show bipartisan support for AI safety measures
- Ipsos found that Biden's executive order on AI received strong bipartisan support, with more than 70% of Democrats and Republicans supporting the development of new standards for AI safety.
Many Americans are self-studying AI — and want to learn more
- Only 3% of Americans consider themselves to be experts on AI, with 29% saying they have some knowledge. But 55% say they would be interested in learning how to use generative AI tools — highlighting the need for education and training in this space.

Americans' attitudes on AI going into 2025

At the end of 2024, nearly half of Americans agreed that the government should have a major role in oversight of AI. Meanwhile, while only one in four think AI will create more jobs than it replaces. (Read more.)

Fewer think AI is coming for white collar jobs

There is is a big jump in people thinking that white collar, blue collar, and their own job prospects will get better thanks to AI, according to the latest round of the Ipsos Consumer Tracker. And perhaps counterintuitively, more now think AI will lead to an improvement in energy use, despite more and more coverage of how energy-demanding it is to run AI systems. (Read more.)

A perceived lack of benefits is a major barrier to adoption of AI tools

One in three people (33%) say they don't think they have any need for AI tools, and another 19% say they don't see the benefit of using them, according to the Ipsos Consumer Tracker. All of this points to a marketing challenge for tech companies – though likely a solvable one with the right creative and media. (Read more.)

The forms of creativity people think AI will help — and harm

History often hints at the future. In the past, tech disruptions led to more jobs being created than lost. There aren’t nearly as many darkroom techs as there used to be, but there are plenty of people creating photography. The thing here is that AI is disrupting so many creative (and less creative) professions all at once. That’s the part that will be hard to forecast. (Read more.)

The AI-driven shifts Americans find most worrying

The democratization of creative tools is revolutionizing the landscape of creativity — but also prompting new fears around the devaluing of human ingenuity. (Read more.)

Many are growing wary of AI-generated content

There’s been a jump in the number of Americans who use negative words to describe AI, compared to the same time last year, according to the Ipsos Consumer Tracker: “Fake” is up 50% to 21%; “Cool” is down from 14% to 10%; “Futuristic” is down, as is “Innovative”; while “Controversial” is up 23% to 34%. (Read more.)

Many are worried about the misuse of AI...

About two in three are worried AI will be misused in policing and law enforcement, or in corporate hiring decisions. Six in ten think there will be too little federal government oversight of AI, while just under half (46%) think that there will be too much federal oversight. (Read more.)

... But many trust AI more than humans

Only 43% of people trust AI tools not to discriminate or show bias towards any group of people, which doesn’t seem great — but they actually trust AI slightly more than they trust humans (only 38%) in this regard. (Read more.)

Some young adults are turning to AI-powered mental health services

 Therapeutic uses of generative AI remain controversial — but recent Ipsos polling suggests that one in five young Americans has experimented with mental health services powered by this technology. (Read more.)

People around the world are worried about rampant AI

60%, on average across 30 countries, are concerned about an AI-based defense system becoming a threat to humanity by breaking free of human control. How real the threat is perceived to be depends on where one lives, with 45% of people in Sweden versus 76% in Indonesia agreeing the threat is real, while 64% of Americans think the threat is real vs. 59% of Canadians. (Read more.)