The Youth Lab

OUI MADAME! NO PULSE. ALL CLOUT.
Gen Z will follow a fake. Just not a fraud.

Meet Jean Phil. A blond bob, a curled moustache, a questionable suit and a habit of shadowboxing through Paris. In under a month he gained over half a million followers across Instagram and TikTok. One small detail: he doesn't exist.

As part of our ongoing AI exploration, THINKHOUSE's in-house AI team had built a virtual influencer before Jean Phil threw his first punch: a fluorescent-orange alien designed to share youth culture insights. Guests at our 2025 Youth Culture Uncovered event got a sneak peek. Getting the tech right was one challenge. Deciding what he should look like was the bigger one, and we made him an alien on purpose. A character nobody could mistake for a real person felt more honest and more memorable, than one designed to pass as human, and the past year has only backed that up.

Youth research firm YPulse crowned Jean Phil Gen Z's new favourite Frenchman. The twist: YPulse's own data shows 70% of young Europeans wouldn't want to follow someone computer-generated. In this edition of 52INSIGHTS, we explore why Gen Z embraces fictional personalities but rejects AI slop, and what that means for influence.

GEN Z SAYS IT WON'T FOLLOW AN AI. THEN IT MET JEAN PHIL.

Jean Phil's first Instagram post went up on 17 September, and BBC News reported his videos have since pulled in tens of millions of views. Within weeks, his face was turning up in other people's videos, and the internet was full of tutorials on how to build your own.

Now look at what Gen Z says about AI. In an Attest survey of 1,000 US 18 to 27 year olds, 41% actively dislike AI-generated content, blaming "AI slop" for lowering quality. Another 31% are wary because it's getting harder to tell what's real. Sprout Social found half of Gen Z have already blocked, muted or unfollowed a brand or creator over content that felt like AI slop.

So why the exception? The same Attest research found 67% of Gen Z engage with comedy and memes, and Jean Phil is pure comedy: one ridiculous character, recognisable in a single frame, doing the same absurd thing in new places. It's part of a wider shift the BBC's Top Comment explored, as AI influencers move away from flawless virtual models towards exaggerated characters.

This aligns with the playbook from our Youth Culture Uncovered 2025 report, where we found three forces driving algorithmic content: Entertainment, Education and Emotion. AI hasn't changed what works. What's changed is the format and ease through which it's delivered. Virtual influencers like Lil Miquela once took a specialist team. Now all it takes is one creator.

Gen Z will watch an AI character. It won't watch AI filler.

THREE THINGS JEAN PHIL SHOWS US ABOUT AI INFLUENCE

1. Character beats realism. Jean Phil works like a good sitcom character: you know exactly what he'll do, and it's funny when he does it. The copycat recipe now circulating has two steps. Build a character sheet, then swap that character into existing viral videos. That's how fast a format spreads when one person can produce it. It's also where the problems start.

2. The joke stops being funny when viewers feel duped. Plenty of people didn't realise Jean Phil was AI, and for some, finding out stung.

"I never get tricked by AI. I was tricked by AI."
Mark Luna, Creative Technologist, THINKHOUSE

His bio linked to a meme coin, which Dexerto reported was valued at over $8 million at its peak. AI literacy educator Jeremy Carrasco called it a “pump-and-dump” scheme. Then there's the copycat recipe. Swapping a character into someone else's viral video is what Top Comment calls AI reskinning, and it's already hitting real creators. Paige Davis found one of her videos reposted with an AI face over hers. When people reported it, the platform said it didn't break community standards. A real person does the performance. The fake face gets the views.

3. When everyone's in on it, fiction can become a real business. Let’s compare this to Chloe vs History. Her creator, British history fan Jonathan Laramy, told Think Like a Creator that she gained 500,000 Instagram followers in 30 days. She vlogs from Tudor London, ancient Rome and the Titanic. Nobody watching thinks it's real footage, and that's the point.

The comments are largely positive. Viewers said they were learning more history from Chloe than at school, many call it "AI for good", and some teachers say they've started showing the videos in class. Laramy and his studio, Majestic Studios, build her scenes from archival photographs, engravings and archaeological records, with historian review. The studio is pitching to museums, educators and broadcasters.

"Technology can generate images. It cannot decide what matters."
Majestic Studios

Like Jean Phil, she's already being copied by other "time traveller" accounts. The difference is what's being built around her: AI film company Utopai Studios points to her as a showcase for its tools, and talent agency Underscore Talent represents her. Laramy is open about the flaws, too, admitting the AI still makes mistakes, such as putting modern road markings on the streets of Victorian London. Jean Phil got attention. Chloe is building trust.

GEN Z DOESN'T HATE AI. IT HATES BEING FOOLED.

Just over two weeks before Jean Phil's first post, Instagram drew a line. On 31 August, Instagram renamed its "AI creator" label to "AI-generated profile" and said it would limit the reach of unlabelled accounts featuring an AI-generated person. Its reasoning was that people dislike a profile that seems human, only to find out later it was made with AI. Chloe's profile now carries the label.

Regulators got there first. Since 2 August, Article 50 of the EU AI Act has required deepfake content to be clearly disclosed. Openly creative, satirical or fictional work still has to disclose, but can do it in a way that doesn't spoil the piece.

The research shows why it matters. A 2026 meta-analysis of 210 experimental studies found virtual influencers match humans on engagement but generally trail them on credibility and on people's intention to act. IAB Ireland and RED C found 80% of Irish adults want significant AI use disclosed, and just 17% think sponsored creator content feels authentic. A Guardian investigation found brands using AI-generated people as fake customers to sell their products. A fake Frenchman for laughs is fine. A fake customer selling you something isn’t.

Back to that 70% of young Europeans who say they wouldn’t follow someone computer-generated. Personas like Jean Phil suggest the objection isn’t to fiction itself, but to being fooled. The youth view, we predict, is that AI labels will stop feeling like a warning and become part of the joke. Gen Z will follow a character it knows is fake, and turn fast on one that pretends otherwise.

When anyone can manufacture a personality, credibility becomes the harder thing to manufacture.

WHAT DOES THIS MEAN FOR BRANDS?

Gen Z has drawn its line, and platforms and regulators are starting to follow suit. For brands, that makes the opportunity clearer.

"Young people will happily laugh along with AI content, but the moment they feel like a brand is trying to dupe them, the trust is gone. For brands, the creative opportunity is real, as long as honesty is built in from the start. To them, the use of AI is not always a problem in itself, it’s the intention behind it"
Aoife Brady, PR, Social & Content Manager, THINKHOUSE

Start with character, not technology. Jean Phil and Chloe work because they're instantly recognisable and do one thing well. Treat a fictional personality as owned IP with a format, a world and a purpose, and judge it on trust and conversion rather than views. Put human oversight on accuracy, permissions and credit.

Then label it. Disclosure is now a platform rule and, in the EU, a legal one. Audiences can love a character they know is fictional. A fake customer, a hidden coin or a borrowed performance will be found out, and Gen Z won't give it a second chance.