What the report actually found
On October 20, 2026, YouTube's Culture & Trends team published a report called Human Creativity and AI, built on survey research conducted with NRG. The survey covered nearly 700 people aged 14 to 44 who had posted a video online in the past 12 months. The topline finding: 72% of these uploaders say they already use AI on videos or images in some capacity, far more than the public conversation about AI skepticism would suggest.
The more interesting detail is how they're using it. The report found that the most popular use of generative AI tools among surveyed posters wasn't automation or faceless production. More than two-thirds of people who've posted AI videos said they featured themselves in that content. And 61% of surveyed video posters said they're specifically interested in using AI to insert themselves into content, essentially using the tools to direct their own lives rather than hide from the camera.
Why this matters for the faceless-channel assumption
A lot of short-form strategy over the past two years has rested on a simple premise: generative AI lets you build a channel without ever appearing on camera, so you should. Faceless automation has been pitched as the main value of AI video tools, letting creators scale output without the friction of filming themselves.
This report complicates that premise. According to the coverage summarizing the findings, the dominant mode isn't faceless automation, it's personal, emotionally legible content where the creator's identity is the point. That doesn't mean faceless content stops working or disappears. But it does mean the audience YouTube surveyed, people aged 14 to 44 who are actively posting video, aren't treating AI primarily as a way to remove themselves from the frame. They're treating it as a way to put themselves into frames they couldn't otherwise reach.
One example cited in the report is the hug your younger self trend, where people use generative AI to create an emotional scene with a younger version of themselves. That's not utility in the sense of saving editing time. It's identity work, using the tool to explore something personal and specific to the creator.
The gap between skepticism and behavior
The report also surfaces a tension worth naming plainly: survey respondents express skepticism about generative video as a category, yet 72% of uploaders are already using AI tools on their content. The brief doesn't fully explain this gap, and it's worth being honest about that. It could mean people are skeptical of AI video in the abstract (deepfakes, low-effort spam, generic slop) while still finding specific personal uses valuable. It could also reflect normal survey behavior where stated attitudes lag actual habits.
What is clear from the available data is that skepticism hasn't stopped adoption. For creators, that's a more useful signal than the skepticism headline alone. The question isn't whether your audience distrusts AI video in general, it's whether your specific use of it reads as personal and legible or as generic and disposable.
What this suggests for series built with AI
If identity, not efficiency, is what's resonating with this 14-to-44 audience, that has a direct implication for anyone building an AI-assisted series: a fully anonymous cast of AI characters may be leaving engagement on the table that a recognizable human throughline could capture.
This doesn't require putting your actual face on camera for every shot. The report's framing is about inserting a self into content, not necessarily live-action self-recording. A consistent avatar, a stylized likeness, or a persona that's recognizably tied to a real creator can serve the same function the survey describes, letting viewers track an identity across a story rather than watching an ensemble of characters with no anchor.
Practically, this suggests a hybrid structure for AI-generated series: keep the efficiency of a consistent AI cast and world-building, but anchor the series around one identity-driven protagonist that viewers can recognize and follow, episode to episode, the way they would a host or a recurring character played by a real person.
- Give your series one consistent, identifiable protagonist rather than a rotating anonymous ensemble
- Use AI tools for world-building and supporting cast, but let a recognizable persona or avatar carry emotional through-lines
- Borrow from formats like hug your younger self: personal, specific, emotionally legible scenarios outperform generic utility content
- Don't assume skepticism about AI video means your audience won't accept AI tools, the data suggests the opposite when identity is central
Where to go from here
The report is a single survey snapshot, nearly 700 people in one age bracket, commissioned by a platform with its own interest in encouraging AI adoption. It shouldn't be read as proof that faceless content is obsolete. But it is a clear data point against the assumption that creators and viewers primarily want AI video to remove the human from the equation.
For creators weighing how to structure an AI-assisted series, the practical move is to test an identity-anchored format alongside any faceless or ensemble-based approach you're already running, and compare retention and comments rather than assuming one model wins by default. Tools like Episodo that support both consistent AI casts and a recurring, recognizable protagonist make that kind of hybrid test straightforward to set up without starting from scratch.