UCLA's Gen Z 'doom' finding is about AI stories, not AI-made video

A 13-year-old boy in Florida was asked what kind of AI he wanted to see on screen, and his answer reads like a development note: "I'd like to see AI portrayed in a balanced and realistic way, showing both its benefits and potential risks." He wanted stories about AI helping people solve problems and learn things, while asking hard questions about trust, privacy, creativity and human control.

His answer sits inside "Rewrite This!", the fifth annual Teens and Screens report from the Center for Scholars & Storytellers at UCLA, published on Wednesday 21 October. It surveyed 1,537 Americans aged 10 to 24, and its AI numbers spread fast. IndieWire's version said young people feel doomed watching AI-generated content, and that the feeling grows with age. The second half of that holds up. The first half describes a question the survey never asked.

What the survey actually asked

Respondents were asked to complete one sentence: "Watching a story with AI in it makes me feel..." They picked one of five answers. Overall, 28.2% chose fascinated, 25.6% excited, 22.0% doomed, 10.1% scared and 14.1% something else. That is where the "over 30% feel doomed or scared" figure comes from (22.0 plus 10.1).

The report's own text calls this a question about "portrayals of AI in the media", and UCLA's release talks about "stories about AI". GamesBeat's Dean Takahashi, who interviewed co-author Yalda Uhls, spelled it out: the poll asked about watching a story with AI as the subject matter, and it did not ask how respondents felt about something made with AI. Doom, he reported, meant the idea that AI is going to kill the human race. So the 22% is a reaction to the killer-machine plot, not a verdict on AI video, AI shorts or anything a filmmaker generated.

Only one question in the study treats AI as a medium. Asked which kinds of entertainment media can tell a story, 41.3% ticked AI. That put it sixth of eight, behind YouTube at 72.9%, TV at 70.3%, movies at 70.2%, games at 68.5% and TikTok at 64.2%. Nobody in the survey was asked whether they would watch, share or pay for something made with AI.

The age split is real, with a catch

The age pattern is the strongest thing in the data. Among 10 to 13 year olds, 9.7% picked doomed and 44.4% picked excited. Among 14 to 18 year olds, doom reached 25.1% and excitement fell to 21.2%. Among 19 to 24 year olds, doom was 27.4% and excitement 17.0%. The youngest viewers are the most open to AI on screen, by a wide margin.

The report heads that section "Fascinated at 12, Doomed by 19", and that overstates its own chart. The jump happens at 14. The two older bands sit 2.3 points apart on doom, which is well inside the uncertainty of a sample this size. The survey also broke out only doom and excitement by age, so the claim that doom becomes the oldest group's dominant emotion is proven against excitement and nothing else. The age bands do not map neatly onto generations either: by the usual 2010 cutoff, many 14 and 15 year olds in the middle band are Gen Alpha too.

A few numbers in circulation are simply wrong. UCLA's release gives 44.2% of the youngest group as "fascinated", where the report's chart labels that line excitement at 44.4%. The release has 17.7% for the oldest group's excitement, where the chart says 17.0%. IndieWire describes the top band as 18 to 24, but 18 year olds sit in the 14 to 18 band.

Who was asked, and who paid

The sample came from an opt-in Qualtrics panel with quotas for gender, age, race and region, surveyed between Sunday 16 August and Tuesday 1 September. That is not a random sample, and the report publishes no margin of error. Each age band holds roughly 400 to 630 people. The report says its open-ended answers were coded using Claude and ChatGPT, so a study about how young people feel about AI had its written answers sorted by AI.

The report thanks Roblox, Disney, YouTube, Hopelab, the Nielsen Foundation and Warner Bros. Television for supporting what it calls independently conducted research. Nothing in the report suggests the funders shaped the AI question, but readers should know who they are. One of them, YouTube, published its own study the day before, run with the research firm NRG. It surveyed nearly 700 Americans aged 14 to 44 who had posted a video in the past year and found 72% use AI on their videos or images. That is a platform measuring makers, not viewers, so it does not cancel out the UCLA numbers, but it shows how far the supply side has already moved.

For a probability-sample benchmark, the report itself cites Pew Research Center's February study of 1,458 US teens aged 13 to 17, run through Ipsos KnowledgePanel with a margin of error of 3.3 points. Pew found more teens expected AI to be good for them than bad, and about four in ten had used chatbots to create or edit images or video.

Where the opening is for AI filmmakers

Read correctly, the study is a brief for people who make things with these tools. The doom in it attaches to a story genre, and the people answering asked for something better than that genre.

• Stories about AI have an audience that wants nuance. The respondents asked for AI shown with its promise and its risks, and AI filmmakers who write about the technology with that balance are writing straight into a gap the data names.
• The youngest viewers are the warmest. Only 9.7% of 10 to 13 year olds picked doom, and that group turns 18 within the next five to eight years.
• AI already registers as a storytelling medium for 41.3% of this audience, about two thirds of TikTok's score, which makes it a baseline to beat next year.
• Story counts for more than method: the same respondents preferred unknown actors (39.8%) to famous ones (38.2%) or influencers (16.8%), and asked first for hopeful, relatable stories.

The teens also named where they already see AI in entertainment: AI fruit videos such as "Fruit Love Island", the "tung tung tung sahur" brainrot memes, AI soap operas on TikTok and vertical-drama apps such as DramaBox and ReelShort. The report gives no counts for those answers, so they show familiarity, not popularity.

Our take: this study does not show a generation rejecting AI video. It shows older teens tired of the AI apocalypse story and younger ones curious about AI on screen. The question that matters for AI filmmakers, whether this audience will choose AI-made work, still has not been asked cleanly.

Timeline

• Date: Thu 25 Sep to Thu 9 Oct 2025. Event: Pew fieldwork, 1,458 US teens aged 13 to 17; Source: Pew Research Center methods
• Date: Tue 24 Feb 2026. Event: Pew publishes "How Teens Use and View AI", later cited by the UCLA report; Source: Pew Research Center
• Date: Sun 16 Aug to Tue 1 Sep 2026. Event: Teens and Screens fieldwork, 1,537 respondents aged 10 to 24, Qualtrics panel; Source: Report methods section
• Date: Tue 20 Oct 2026. Event: YouTube Culture & Trends publishes its maker survey with NRG; Source: YouTube Blog
• Date: Wed 21 Oct 2026, 13:30 UTC. Event: UCLA release and GamesBeat interview with Uhls published; Source: UCLA Newsroom; GamesBeat
• Date: Wed 21 Oct 2026, 16:00 UTC. Event: IndieWire story frames the finding as AI-generated content; Source: IndieWire
• Date: Thu 22 Oct 2026. Event: Sixth annual Teens and Screens Summit, invitation only; Source: UCLA Newsroom

Sources and what is unconfirmed: the full report PDF, "Rewrite This! How Gen Z and Gen Alpha Are Redefining Modern Storytelling" (Rivas-Lara and Uhls), including its methods section and chart pages; the Center for Scholars & Storytellers report page; the UCLA Newsroom release (21 Oct); GamesBeat (Dean Takahashi, 21 Oct); IndieWire (Brian Welk, 21 Oct); YouTube's blog post on its maker survey (20 Oct) and Social Media Today's summary of it; Pew Research Center (24 Feb). The line that the poll did not ask about AI-made content is GamesBeat's reporting after its interview with Uhls, not a direct quote, though the report's wording supports it. Per-age shares for fascinated, scared and other are not published. Band sizes are our arithmetic from the published shares. We did not read YouTube's full report, and we have not contacted the Center.