How do Americans feel about Generative AI? It’s complicated.

KEY Findings:

1. In polling conducted six months after ChatGPT’s release, one in six (16%) Americans say they have used some type of generative AI system.
2. One in five (20%) say they have used an AI chat or text program in the past month.
3. Americans want businesses to regulate AI, though trust with these business is tentative. Just like social media, Americans want to see AI regulated, but not outright banned.
4. Political polarization is not an issue for AI yet, but the potential for it happening is there.

Technology and society are intertwined, and as technology has further entrenched itself into our everyday lives, politics has become infused into the public’s perception and experience of new platforms and products.

Now, after waves of innovation, artificial intelligence (AI) is the next development set to disrupt businesses, governments, and culture. With this fast-moving field, we find that most Americans are just tuning into this conversation, are not quite sure what to think yet and do not necessarily see it as a disruptive force, despite how some executives and politicians talk about artificial intelligence. Yet, most want to see AI regulated by the businesses developing these products—though that does not mean they do not see serious risks involved in going down this regulatory route.

This is a crucial time for executives and government officials using, developing, and regulating AI to maintain public trust. Missteps now will linger within the industry.

AI adoption has been remarkably quick, but still relatively small.

Relative to other technologies, Americans have been quick to adopt AI. In just around two months, ChatGPT quickly reached 100 million users worldwide, doing so in a fraction of the time it took other viral apps like TikTok, Instagram, or Facebook, surpassed only recently by Threads. But despite the rapid growth in the public eye, it is important for policymakers and business leaders to note that only a fraction of Americans say they are actually using these technologies.

In polling conducted six months after the release of ChatGPT, one in six (16%) Americans say they have used some type of generative AI system, like the text-based app ChatGPT or the image generating app DALL-E. Additionally, one in five (20%) say they have used an AI chat or text program in the past month. By comparison, use of AI has already more than doubled the use rates of other tech trends, like cryptocurrency or metaverse systems, even if it has not reached the prevalence of online search.

How Americans are using AI varies. Among the Americans that say they used a generative AI system, like ChatGPT, Bard, or DALL-E, at some point, the most common use has been to create images, to create text for their jobs, or write computer code. One in ten Americans that have used AI have used it for schoolwork. Younger Americans are also more likely to have used AI than older Americans.

In a short time, AI elbowed its way into the spotlight. However, executives should note that most of the public is still largely unfamiliar with AI, even as some people have been quick to try out things like ChatGPT.

Americans have mixed feelings about AI. Most do not see AI as disruptive.

As more Americans are becoming familiar with AI, it is crucial for executives to understand that early experiences can strongly shape sentiments. Right now, a plurality say they are curious, but even more say they are uninterested, scared or excited. Given the relative newness of the technology and the high degree of uncertainty in its future, Americans are not entirely sure how to feel about the technology yet.

Most Americans also see AI as being just another piece of technology among many, as opposed to feeling AI could fundamentally change American society, the position of many AI evangelists.

Whether those changes are good or bad depends on who you ask. Younger Americans are more likely to see AI optimistically. The majority of adults under 30 (58%), for example, feel AI will generally have a good impact on the lives of average Americans, while less than half of Americans over 65 (42%) think the same. 

As business leaders introduce AI into their work streams and public facing products, these executives should consider how the public is not necessarily viewing AI as transformative technology yet, but opinion on AI is malleable and subject-to-change; what is acceptable with the public now might not be the case later.

Americans want businesses to regulate AI, though few have significant trust in business to do so.

As leaders grapple with how to grow the use of artificial intelligence in American society, they should understand that Americans do not have a clear consensus on how they would like AI to be regulated, even as many fear the repercussions the technology could bring.

Americans are near split when it comes to what entity – business or government – should regulate AI. Slightly more Americans think that the responsibility to develop AI to minimize harm should lie with the company developing AI (53%), while 44% say it is the government’s responsibility to set the rules and limit the risks around AI.

However, despite the public slightly preferring business to self-regulate, the vast majority of Americans (75%) distrust companies developing AI to do so with the public’s best interest in mind. Furthermore, more view the unchecked development of AI driving disinformation and economic chaos (75%) as a far greater threat than government intervention slowing down the progression of U.S.-based AI technology (21%).

Decision-makers should understand that a majority of the public currently prefer a lighter-touch regulatory system, but that is not a carte blanche. The public does not want this to come at the expense of disinformation or economic turmoil, and sentiment is likely to change abruptly in the wake of a crisis.

Just like social media, Americans want some AI regulation, but no bans.

Government officials and business leaders can look to other technologies to understand how the public is thinking about regulating and approaching artificial intelligence. Americans’ views on AI regulation hold parallels to social media, with Americans supporting some regulations but no outright bans or limits.

Majorities want to see social media regulated by companies over the government. Most Americans feel there should be some common-sense guardrails put up for AI and social media, such as requiring that companies state when AI generated content is being used and regulating ad campaigns on social like they are on TV and radio. But, in both cases, most Americans do not support government regulation outright banning or limiting either AI or social media content.

Polarization is not an issue for AI yet, but it is a big concern.

The tech industry is rife with partisan division. But AI has not (yet) become a partisan flashpoint to the same degree as social media. It is important for leaders to remember that the more proximate the tech business model is to polarization, the greater the reputational and regulatory risks. AI will not be an exception.

Republicans already view AI less favorably than Democrats or independents. Though, on the whole, partisanship is not as pronounced around artificial intelligence as it is among other tech products, like social media.

What’s Next

Generative AI is forcing business leaders and regulators to reimagine the role technology plays in our lives, leaving the tech industry, policymakers, and Americans writ large scrambling for what is to come.

Right now, no one has answers. Even the questions we should be asking are unclear. Even so, it is vital that decisions be made with the public’s opinion in mind. By keeping these three points in mind, executives can better read the room on artificial intelligence—and make better decisions in the future.

1. Americans are feeling a wide range of emotions towards AI. Most do not see AI as disruptive.
2. The public would rather have companies developing AI to regulate AI themselves rather than the government—but Americans hold reservations with private-sector regulation, too.
3. AI might not be an exception in our partisan times. The closer the technology or business model is to polarization, the greater the reputational and regulatory risks.