WASHINGTON, D.C. -- Nearly all Americans use products that involve artificial intelligence (AI) features, but nearly two-thirds (64%) don’t realize it, according to a new study conducted in partnership with Telescope, an organization that works toward the responsible advancement of technology. While some Americans are optimistic about how AI will contribute to society in the future, many view its potential impact negatively. Most U.S. adults, regardless of political party, believe the government is at least partially responsible for addressing the potential harms and risks associated with AI. In addition, majorities say business leaders also bear responsibility for how AI is used.
These findings are based on a nationally representative web study of nearly 4,000 U.S. adults, conducted Nov. 26-Dec. 4, 2024. For the purposes of this research, artificial intelligence (AI) is defined as “computer systems able to perform tasks that normally require human intelligence.”
Most U.S. Adults Don’t Realize Usage of AI-Enabled Products
When asked whether they have used some kind of AI-enabled product in the past seven days, 36% of Americans report that they have, while 50% say they have not and 14% say they aren’t sure.
However, when asked about their usage of six common AI-enabled products (personal virtual assistants, navigation apps, weather forecasting apps or websites, social media platforms, streaming services, or online shopping apps or websites), 99% of U.S. adults report using at least one of these in the past week, with 83% saying they have used at least four.
Half of those initially saying they have not used an AI-enabled product in the past seven days report using five or six of the listed products in the past week, and another 37% report using three or four.
Over eight in 10 Americans report using five common AI-enabled products within the past seven days -- weather forecasting apps or websites (87%), streaming services (83%), online shopping websites or apps (82%), social media platforms (81%), and navigation apps or websites (81%).
In addition, half of Americans (50%) say they have used a personal virtual assistant such as Amazon Alexa, Google Assistant, Apple Siri or Microsoft Cortana in the past week. Those initially saying they have used an AI-enabled product are considerably more likely to report interacting with a personal virtual assistant (71%). However, more than a third of those saying they have not used an AI-enabled product also report using a personal virtual assistant in the past week (35%), as do nearly half of those who are unsure whether they’ve used an AI-enabled product (47%).
Adults aged 18 to 29 (46%) are significantly more likely than those aged 65 and older (23%) to report using AI weekly. Perceived usage of AI is also higher among Americans with a college or graduate degree (47%, vs. 31% among those without a degree), those earning at least $120,000 in annual household income (45%, vs. 38% making $60,000 to less than $120,000 and 30% making less than $60,000), and those living in large cities (44%, vs. 38% living in suburbs and 31% living in rural areas).
Most Americans View AI’s Potential Impact Negatively
Overall, more Americans have negative than positive views of AI’s potential impact on various aspects of society. Majorities expect that AI will have a negative impact on the spread of false information (72% “somewhat” or “very” negative), social connections among people (64%), Americans’ job opportunities (60%) and U.S. national security (56%) and in the next five years. AI’s future impact on detecting nonviolent criminal offenses is about equally split. The only area asked about that receives more positive than negative views is AI’s impact on medical diagnosis and treatment.
In general, Americans who report using AI-enabled products in the past week are more likely than those who do not to have positive views of AI’s potential impact. The one exception is for the spread of false information, for which there is no meaningful difference by reported AI usage.
Americans who report using a higher number of common AI-enabled products have more positive views of the potential for AI technology in medical diagnosis and treatment: 63% of those who report using five or six of the listed products have somewhat or very positive perceptions of medical applications for AI, compared with 58% of those who report using three or four products and 54% who reporting using fewer than three.
In contrast, using more AI-enabled products is associated with more negative perceptions of how AI will affect the spread of false information in the next five years. While 64% of Americans who use three or fewer products feel AI technology will have a somewhat or very negative impact on the spread of false information, 71% of those who use three or four of the products and 75% of those who use five or six products have a negative view.
Younger adults, those aged 18 to 29, are more likely than older adults to have negative perceptions of the impact of AI technology, especially on job opportunities. Seventy-eight percent of younger adults say they think AI will have a very or somewhat negative impact on Americans’ opportunities to get good jobs in the next five years, compared with 66% of adults aged 30 to 39, 58% of adults aged 40 to 64, and just 45% of adults aged 65 and older.
Overall, there are minimal differences in attitudes by political affiliation. The largest gaps are seen for the effect of AI on national security (59% vs. 50%) and the spread of false information (78% vs. 66%), with Democrats reporting more negative attitudes than Republicans in both cases.
Americans Aligned on Responsibility for Reducing Potential AI Harms
Most Americans say the government is at least partially responsible for reducing the potential harms caused by AI. A nearly unanimous 96% say the government is mostly or partly responsible for reducing AI-related national security threats, including 62% who say the government, as opposed to business, is primarily responsible.
Large majorities of Americans also believe the government is at least partially responsible for addressing potential AI problems related to the spread of false information (88%), personal data privacy violations (82%), unauthorized use of individuals’ likeness (81%) and job displacement (68%).
Perceptions of government responsibility are similar among Republicans and Democrats, although Republicans are slightly more likely to say business should take primary responsibility for negative effects on workers displaced from their jobs by AI technology. For Republicans and Democrats alike, job displacement is the one area in which Americans see businesses as having considerably more responsibility than the government: 84% say businesses are mostly (32%) or equally (52%) responsible, while 68% say the government is mostly (16%) or equally (52%) responsible.
Although perceived responsibility for businesses is highest for job displacement effects, when it comes to other areas of potential AI-related harm, businesses are not off the hook in Americans’ eyes. Over six in 10 Americans say the government and businesses are equally responsible for addressing problems related to the spread of false information (68%), personal data privacy violations (63%) and unauthorized use of individuals’ likeness (62%). A slim 52% majority also believe the government and businesses share responsibility for reducing harm to workers who lose their jobs because of AI technology.
Demographic differences in perceptions of responsibility are generally small and vary across different types of potential AI-related problems.
About Half of Americans Feel High Personal Accountability for Responsible AI Use
When asked how much personal accountability they should take for using AI technology responsibly, 55% of Americans believe they should take “a lot” of personal accountability, and 32% say they should take “some.”
Views of personal accountability for responsible AI use are not any higher among respondents who report using multiple AI-enabled products each week -- 54% of those who use five or six of the listed products say they should take a lot of responsibility, with one in seven (14%) saying they feel they should take very little to none.
Among younger adults, who report the highest level of AI-product usage and the most negative views of AI’s impact in the next five years, perceptions of personal accountability are even lower. Nearly a quarter (23%) say they should take very little to no personal accountability for using AI technology responsibly, over twice the rate of those aged 40 and older.
Implications
As AI technology rapidly advances in the U.S. and abroad, products that Americans use frequently in their daily lives are increasingly adopting AI to improve their efficiency, accuracy and personalization. However, it is not always evident to users that a product uses AI features.
This research confirms that nearly all Americans report using products with AI features, but most do not realize that common apps, sites and web tools such as weather forecasting, navigation and online shopping use AI technology. Many also do not recognize that virtual personal assistants embedded in smartphones, tablets and other devices are a form of AI-enabled technology.
Despite the ubiquity of AI technology in daily life, many Americans hold negative views of its potential influence on jobs, social connections and security in the next five years and want to see the U.S. government and businesses take a role in reducing the potential harms AI may cause in the future.
At the same time, only about half of U.S. adults feel a high level of personal accountability for their own use of AI. These results suggest that many Americans may lack awareness about how their own behavior could shape AI’s impact on U.S. society, offering an opportunity to provide guidance to help Americans incorporate AI into their lives responsibly.
To stay up to date with the latest Gallup News insights and updates, follow us on X @Gallup.
Learn more about how the Gallup Panel works.
