What Teenagers Really Think About AI

American teenagers believe addressing the potential risks of artificial intelligence should be a top priority for lawmakers, according to a new poll that provides the first in-depth look into young people’s concerns about the technology.

The poll, carried out by youth-led advocacy group the Center for Youth and AI and polling organization YouGov, and shared exclusively with TIME, reveals a level of concern that rivals long standing issues like social inequality and climate change.

The poll of 1,017 U.S. teens aged 13 to 18 was carried out in late July and early August, and found that 80% of respondents believed it was “extremely” or “somewhat” important for lawmakers to address the risks posed by AI, falling just below healthcare access and affordability in terms of issues they said were a top priority. That surpassed social inequality (78%) and climate change (77%).

Although the sample size is fairly small, it gives an insight into how young people are thinking about technology, which has often been embedded in their lives from an early age. “I think our generation has a unique perspective,” says Saheb Gulati, 17, who co-founded the Center for Youth and AI with Jason Hausenloy, 19. “That's not in spite of our age, but specifically because of it.” Because today’s teens have grown up using digital technology, Gulati says, they have confronted questions of its societal impacts more than older generations.

Read More: 5 Steps Parents Should Take to Help Kids Use AI Safely

While there has been more research about how young people are using AI, for example to help or cheat with schoolwork, says Rachel Hanebutt, assistant professor at Georgetown University’s Thrive Center, who helped advise on the polls’ analysis, “Some of those can feel a little superficial and not as focused on what teens and young people think about AI and its role in their future, which I think is where this brings a lot of value."

The findings show that nearly half of the respondents use ChatGPT or similar tools several times per week, aligning with another recent poll that suggests teens have embraced AI faster than their parents. But being early-adopters hasn’t translated into “full-throated optimism,” Hausenloy says.

Teens are at the heart of many debates over artificial intelligence, from the impact of social media algorithms to deep fake nudes. This week it emerged that a mother is suing Character.ai and Google after her son allegedly became obsessed with the chatbot before committing suicide. Yet, "ages 13 to 18 are not always represented in full political polls," says Hanebutt. This research gives adults a better understanding of “what teens and young people think about AI and its role in their future," rather than just how they're using it, Hanebutt says. She notes the need for future polling that explores how teenagers expect lawmakers to act on the issue.

Read More: Column: How AI-Powered Tech Can Harm Children

While the poll didn't ask about specific policies, it does offer insight into the AI risks of concern to the greatest number of teens, with immediate threats topping the list. AI-generated misinformation worried the largest proportion of respondents at 59%, closely followed by deepfakes at 58%. However, the poll reveals that many young people are also concerned about the technology’s longer term trajectory, with 47% saying they are concerned about the potential for advanced autonomous AI to escape human control. Nearly two-thirds said they consider the implications of AI when planning their career.

Hausenloy says that the poll is just the first step in the Center for Youth and AI’s ambitions to ensure young people are “represented, prepared and protected” when it comes to AI.

The poll suggests that, despite concerns in other areas, young people are generally supportive of AI-generated creative works. More than half of respondents (57%) were in favor of AI-generated art, film, and music, while only 26% opposed it. Less than a third of teens were concerned about AI copyright violations.

On the question of befriending AI, respondents were divided, with 46% saying AI companionship is acceptable compared with 44% saying it’s unacceptable. On the other hand, most teens (68%) opposed romantic relationships with AI, compared to only 24% who find them acceptable.

Read more: AI-Human Romances Are Flourishing—And This Is Just the Beginning

“This is the first and most comprehensive view on youth attitudes on AI I have ever seen,” says Sneha Revanur, founder and president of Encode Justice, a youth-led, AI-focused civil-society group, which helped advise on the survey questions. Revanur was the youngest participant at a White House roundtable about AI back in July 2023, and more recently the youngest to participate in the 2024 World Economic Forum in Davos.

In the past, she says Encode Justice was speaking on behalf of their generation without hard numbers to back them, but “we'll be coming into future meetings with policymakers armed with this data, and armed with the fact that we do actually have a fair amount of young people who are thinking about these risks.”

Read more: U.S. Voters Value Safe AI Development Over Racing Against China, Poll Shows

She points to the California Senate Bill 1047—which would have required AI companies to implement safety measures to protect the public from potential harms from their technology—as a case where public concerns about the technology were overlooked. “In California, we just saw Governor Gavin Newsom veto a sweeping AI safety bill that was supported by a broad coalition, including our organization, Anthropic, Elon Musk, actors in Hollywood and labor unions,” Revanur says. “That was the first time that we saw this splintering in the narrative that the public doesn't care about AI policy. And I think that this poll is actually just one more crack in that narrative.”