While generative artificial intelligence (AI) has increasingly found itself on the forefront of American consciousness — with marketers eager to take advantage of the new technology and regulators worldwide considering how to regulate it — many consumers remain wary. Most Americans want companies themselves to limit the risks, even if they don’t trust them to do so, according to a new report from Ipsos.
The report, published on May 4, found 53% of Americans believe it is the job of companies developing tools such as ChatGPT to control the risks. However, 75% of those surveyed have little to no trust in AI companies to do so. While companies embrace AI, more Americans view the technology unfavorably (43%) than favorably (39%).
“There is a lot of enthusiasm but also apprehension about the advent of AI,” said Chris Jackson, senior vice president of Ipsos public affairs. “There are a lot of people that are excited about the opportunity, excited about what it can do. But there's also a lot of concern with what's happening with AI.”
The poll was conducted April 20 to 24, 2023, using the probability-based KnowledgePanel. The sample was nationally representative, totaling 1,008 people ages 18 years or older. The sample was also politically diverse and included 286 Republicans, 292 Democrats and 323 independents.
While AI seems to be everywhere, very few Americans have actually engaged with the technology. Only 16% of Americans have used generative AI, with the most common use being to create art, pictures or images (41%). Other popular uses include to create text for a job (29%), generate social media posts (14%), do schoolwork (11%) and create videos (9%). While the tech’s use is varied, an overwhelming majority of Americans have never interacted with the technology.
Additionally, the public’s opinion on the impact AI will have on technology remains varied. While tech companies such as Microsoft, Meta and Google are seemingly ready to go all-in on the technology, only 36% of survey respondents believe AI will fundamentally change American society. Fifty-five percent think AI will be just another piece of technology, while only 6% believe it will have little impact.
With such low uptake and divided opinions on AI’s place in society, the hype around AI is seemingly reminiscent of the scramble surrounding the metaverse, which has lost significant steam in recent months. However, according to Jackson, there are differences between the two that give AI an edge.
“There's already more people who say they've used an AI system, then say they’ve used a metaverse system, so it has certainly grown faster,” said Jackson. “And I think the big question mark is going to be, is there some sort of application of AI that is useful to people.”
(Lack of) responsibility
With generative AI technologies becoming more common, consumers squarely place the responsibility of oversight on companies over the government. A growing mistrust in corporations more broadly could be behind the contradictory finding that only a quarter of Americans trust companies to regulate themselves when it comes to AI.
“The public view of this seems contradictory because what we see is that the public wants the private sector to actually police itself, to regulate itself, not necessarily the government to do that,” said Jackson. “The large majority of Americans don't have a lot of confidence in the private sector companies developing these technologies to do so with the public's interests at heart.”
It’s important to note that while Americans are concerned about AI, they are more concerned with its direct impact on them. Seventy-one percent of those surveyed reported they are concerned with how AI could impact job security and society. Additionally, 76% are concerned about the rise of deepfakes driving misinformation.
While these numbers are high, they lag behind other technological concerns held by Americans. For example, 86% of Americans are concerned about hacking, malware and data breaches, and 78% are concerned about social-media driven radicalization.
“I think there is a practicality to Americans,” said Jackson. “They're sort of looking at the very individual impacts on them as you know, as people and then sort of a second order of bigger picture more abstract concerns.”