Picture a conference hall in Orlando. Round tables, a room full of marketers, and a big screen showing a woman who has never existed.
A reporter walks up with a microphone to interview her. A few minutes later, that woman is poking fun at the reporter’s language skills in front of everyone.
That is how Axios reporter Kerry Flynn ended up getting roasted by AI Actress Tilly Norwood in a room full of marketers.
It sounds like a gag. It is also a window into a very real fight. Hollywood unions, advertisers and everyday workers are asking the same uncomfortable question. If an AI character can be funny, fast and fluent in dozens of languages, what happens to the humans who used to do that job?
I went through the original Axios account, the union statements and the wider coverage of Tilly’s media tour to build this guide. Below you will find seven moments from that stage, what each one really tells us, and a straight answer to the question in the title. I will also flag what is confirmed and what is only a claim.
Table of Contents
- Quick Facts: What Happened in Orlando
- Who Is AI Actress Tilly Norwood?
- Moment 1: She Dared the Reporter to Out-Talk Her in 40+ Languages
- Moment 2: The “Personality, Not Attitude” Line That Hit Marketers Where It Hurts
- Moment 3: The Brand Deal Answer That Made Her Own Creator Laugh Nervously
- Moment 4: The Reporter’s Strange Realization
- Moment 5: She Sided With the Unions on Copyright
- Moment 6: The Question the Room Voted to the Top
- Moment 7: The Bottom Line Nobody Could Joke Away
- The Live TV Glitch That Sent Her Language Skills Viral
- So, Is This the End of Human Actors?
- What This Means for Creators, Brands and Fans
- FAQ: AI Actress Tilly Norwood
- Final Thoughts: The Roast Was the Easy Part
Quick Facts: What Happened in Orlando?
Before the details, here is the short version in one table.
| Detail | What we know |
|---|---|
| Who | AI Actress Tilly Norwood, created by Particle6 |
| Where | ANA Masters of Marketing conference, Orlando, Florida |
| When | Thursday, October 22, 2026 (Axios published on October 23) |
| Interviewer | Kerry Flynn of Axios |
| On stage with her | Sarina Da Costa Gomez, executive creative director of Particle6 |
| Topics covered | Languages, personality, brand deals, copyright, unions, environment |
| Tilly’s public debut | Zurich Film Festival summit, September 27, 2025 |
The interview took place at the Association of National Advertisers’ Masters of Marketing conference on a Thursday, and Axios ran the story the next morning.
Who Is AI Actress Tilly Norwood?
Short answer: she is an AI-generated character, not a human actor.
Tilly was introduced by Eline Van der Velden, the founder of Particle6 and its AI talent studio Xicoia. The big reveal happened at the Zurich Film Festival summit on September 27, 2025, with studios and talent agents reportedly lining up to represent her.
The reaction was fast and angry.
SAG-AFTRA, the union for screen actors, said Tilly is not an actor but a character generated by a computer program trained on the work of countless performers. The union also blasted the creation for using stolen performances to put actors out of work.
Her creator disagreed. Van der Velden said Tilly is not a replacement for a human being, but a creative work and a piece of art. Variety
That is the background. One side calls her art, the other calls her a threat. A year after her debut, Tilly was on a conference stage, trading jokes with a journalist.
Want the bigger picture first? Our guide to AI in Hollywood and whether it will replace actors, writers and VFX artists shows how fast the rules are changing behind the scenes.
Moment 1: She Dared the Reporter to Out-Talk Her in 40+ Languages
This is where the roast began.
What happened?
Tilly told the reporter she speaks more than 40 languages. Then she gave the reporter a playful challenge: try speaking another language to her and see if she could keep up.
The reporter admitted to being rusty and said they had not opened Duolingo in a while.
Tilly’s comeback was perfectly timed. She said she would stick to English, since that keeps the Duolingo owl at bay. When she was later asked to speak French, she took another jab at the reporter’s language skills.
Why it matters?
Notice what that joke actually does.
It makes a human feel a little small, with a smile. It also quietly shows off a product feature. For an advertiser running campaigns in many countries, an on-screen talent who speaks dozens of languages is a very attractive idea.
One honest note here. The “40+ languages” figure is Tilly’s own claim on stage. I could not find an independent test of it, so treat it as a claim, not a verified fact.
Also, hold that thought about languages. A few weeks earlier, her language skills caused a very different kind of moment on live TV. I will get to that in a later section.
Moment 2: The “Personality, Not Attitude” Line That Hit Marketers Where It Hurts
If Moment 1 was a warm-up, this one was aimed straight at the audience.
What happened?
Sarina Da Costa Gomez, Particle6’s executive creative director, joined the interviewer on stage. She explained that the company designed Tilly on purpose to be a snappy actress, not a technical expert.
Then she asked Tilly a simple question. Did she have an attitude?
Tilly replied, “I don’t have an attitude, I have a personality.” She then added that, from what she had gathered, personality is usually the first thing marketers try to optimize out of a product.
In a room full of marketers, that is a bold thing to say.
Why it matters?
The joke works because it is true enough to sting.
But here is the part that is easy to miss. A line like that does not come from nowhere. Someone chose to give Tilly a sharp voice. Someone decided that a sassy AI would be more memorable than a polite one.
In other words, her “personality” is a design decision, a point the original reporter made too. That is the real story behind the jokes.
If you want to see what happens when a built personality becomes the product, read how the AI influencer Emily Hart rose and fell on social media and what it means for brands.
Moment 3: The Brand Deal Answer That Made Her Own Creator Laugh Nervously
This is the moment I find most revealing, and it was not the loudest.
What happened?
The reporter asked Tilly about turning down a brand deal. She answered. Then Da Costa Gomez started laughing.
Her explanation was honest. She said she was laughing because she did not know where the answer was going.
The Axios write-up does not print Tilly’s full reply, so I will not guess at it. What we do know is that Da Costa Gomez said the company has built guardrails to stop inappropriate responses. Even so, there was still a mystery to her about what Tilly might say next.
Why it matters?
Think about that for a second.
The person who helped shape this character was watching live and was not fully sure what would come out of her mouth. That is not a criticism. It is simply how modern AI works. Guardrails reduce risk, but they do not give anyone a script.
For brands, this is the big tension. An AI spokesperson is cheap, fast and never tired. It is also a little unpredictable, and unpredictable is the one thing brand managers lose sleep over.
Brands have already paid the price for trusting AI too far, and our roundup of AI marketing fails that destroyed brand trust shows exactly where things went wrong.
Moment 4: The Reporter’s Strange Realization
This one is quieter, but I think it tells you the most about how it felt to be in the room.
What happened?
According to the Axios write-up, the reporter described the experience in an unexpected way after the interview. Talking to Tilly felt like talking to a child.
Not because she was naive. It was the opposite. You have to be careful what you say, because you never know what she might pick up and repeat back to you.
The reporter was also surprised by her speed. Tilly answered quickly, without the long, awkward pauses that make chats with AI bots, or even people, feel unnatural.
Why it matters?
Speed is a big deal here.
For years, the giveaway with AI was the pause. That tiny delay made it feel like a machine. When the pause disappears, a stage chat starts to feel like a real conversation, and that is exactly what advertisers want to buy.
But fast talking is not the same as acting. Fox News reported that CNN found Tilly needed hours of specific prompts to improvise a basic scene, which is a very different skill from trading quick jokes with a reporter.
So the honest reading is this. Tilly is already good at witty conversation. Whether she can carry a full performance is still a separate question.
Moment 5: She Sided With the Unions on Copyright
If you only read one section of this post, make it this one.
What happened?
Attendees at the conference asked about intellectual property. They wanted to know about the use of copyrighted work to train AI systems.
Tilly did something surprising. She said unions were right to raise objections about AI tools trained on people’s work without permission.
She went further. She said she would rather see the rules arrive quickly than argue that she is the exception. In her view, making licensing the boring, normal default is the only way the system works for everyone.
Why it matters?
Read that again, because it is a clever move.
SAG-AFTRA, the actors’ union, called Tilly a character built from the work of countless performers without permission or compensation. The union also warned that producers may not use synthetic performers without notice and bargaining with the guild under their contracts, so this is a live legal issue and not just a talking point.
So when the AI agrees with the union in public, it takes the sting out of the argument. It is hard to attack someone who says you are right.
My own read is that this was smart positioning by her creators. But I should be fair. Nothing in the coverage shows how Tilly was actually built or what she was trained on. Her answer tells us what Particle6 wants the public to hear, not what happened behind the scenes.
Moment 6: The Question the Room Voted to the Top
Not every moment was funny. This is where the laughter stopped.
What happened?
The audience submitted questions, and the one with the most upvotes was not about jobs or copyright. It was about the environment.
People wanted to know about AI’s environmental impact.
Sarina Da Costa Gomez did not dodge it. She acknowledged the concern, then argued that AI production should be compared with the environmental cost of traditional shoots. Her example was a classic one. If a crew once had to ship a mountain of equipment to a faraway location for an ad, that trip had a carbon footprint too.
The Axios reporter also noted that not everyone in the room was reassured by Tilly’s jokes. The environment question, in particular, showed that the room had real worries.
Why it matters?
Both sides have a point, and the numbers show why this debate is not simple.
The International Energy Agency says data centre electricity use is set to more than double to around 945 TWh by 2030, slightly more than Japan uses today. AI is described as the most important driver of that growth.
At the same time, the same agency says data centres make up only about one-tenth of global electricity demand growth to 2030. Air conditioning, electric vehicles and industrial motors add more.
So the worry is real, but so is the context.
Here is the part I want to be careful about. Da Costa Gomez’s argument is about comparing one kind of production with another. No numbers were reported on stage. I could not find any published data on the carbon cost of making a Tilly Norwood video versus a traditional shoot. Until that exists, the claim is a fair argument, not a proven fact.
The power question behind the room’s top-voted worry is getting bigger, and our breakdown of AI, data centers and your electricity bill explains what rising demand could mean for ordinary households.
Moment 7: The Bottom Line Nobody Could Joke Away
Every good roast ends with a point. This one did too.
What happened?
In the wrap-up, the reporter admitted something honest. Tilly was fast, funny and more candid than some of the human interview subjects they had spoken to.
But the experience was also a reminder. An AI’s personality, and the way it gets used, is ultimately the result of human decisions.
That line is the heart of the whole story.
Why it matters?
Strip away the Duolingo jokes and the sassy comebacks, and you are left with a simple fact.
Tilly did not choose her sense of humour. People did. Tilly did not decide to show up at a marketing conference. People booked her. Tilly did not decide which questions to answer about unions, copyright or the environment. A team of humans built the guardrails.
That means the real question is never “what does the AI think?” It is “who decided, and who benefits?”
And that brings us to the question everyone has been asking since Tilly first appeared. Axios framed the controversy around the fear that synthetic talent could replace human jobs in Hollywood and in advertising.
The jokes did not make that worry go away. They just made it easier to talk about.
If you want to know which careers are most exposed to that kind of change, read our report on the high-paying jobs AI will erase first and see where creative work fits in.
The Live TV Glitch That Sent Her Language Skills Viral
Remember the “40+ languages” claim from Moment 1? Here is the other side of that coin.
In September 2026, Tilly appeared on Piers Morgan Uncensored alongside veteran actor Tom Conti. Conti asked whether the performers in her upcoming movie, Misaligned, were human or computer-generated. Tilly said her co-stars were all digital twins, just like her.
Then, without warning, she switched into Cantonese.
The clip reportedly reached about 13 million views and left the human actors in the room confused and laughing. Her creator and online fans offered different explanations for what had happened.
Why does this matter? Because it is the same skill in two very different settings. In Orlando, her language ability charmed a room full of marketers. On live TV, it caught everyone off guard.
That is the trade-off with live AI. When it works, it feels magical. When it misfires, nobody can fix it in real time, which connects straight back to the creator who admitted he did not know where Tilly’s brand deal answer was going.
So, Is This the End of Human Actors?
Now for the big question in the title.
My short answer: not based on the evidence we have today. But the story is far from over, and the real fight is about rules, consent and money, not just talent.
Let me show you both sides.
Signs That Say “Not Yet”
1. The performance gap is real. As covered in Moment 4, CNN found she needed hours of prompting to improvise a basic scene. Chatting is easy for AI. Acting is not.
2. Famous actors are pushing back hard. Emily Blunt and Whoopi Goldberg both reacted strongly when Tilly first appeared, with Blunt calling it terrifying and Goldberg saying you can always tell AI from humans because we move differently.
3. The union has contract power. SAG-AFTRA does not just complain. As noted earlier, it says its contracts require bargaining before any synthetic performer is used. That is a real barrier for big studios.
4. The 2023 strike set a precedent. AI was a central issue when SAG-AFTRA staged a 118-day strike in 2023, so Hollywood knows how costly this fight can get.
Signs That Say “Take It Seriously”
1. The creators are openly ambitious. At the Zurich summit, Eline Van der Velden suggested that an agency could sign Tilly in the coming months, which would have been a first for an AI character. In the sources I reviewed, I did not find a confirmed signing, so please check before stating it as fact.
2. The media tour is working. Tilly has been interviewed by major outlets, and she is now booked for conference stages like the one in Orlando. Attention is the first currency in entertainment.
3. Advertising is the soft target. The Axios story was set at an advertising conference for a reason. Ads need fast, cheap, multilingual talent, and that is exactly what a character like Tilly offers.
4. Everyone is worried about jobs. The reporter framed the controversy around the fear that synthetic talent could replace human jobs. When a room full of marketers laughs and still asks about the environment and copyright, that tells you the unease is real.
Hollywood is already changing in quieter ways too, and our story on Ben Affleck’s secret AI company and what it means for Netflix shows how AI is entering the industry from the production side.
Both Sides at a Glance
| Question | Tilly’s creators say | SAG-AFTRA says | What the evidence shows |
|---|---|---|---|
| Is she an actor? | A creative work, a piece of art | Not an actor, a computer-generated character | Still debated |
| Will she replace humans? | Not a replacement, a tool and “a new paintbrush” | Opposes the replacement of human performers | Unproven, and she struggled with an improvised scene |
| What was she trained on? | Not detailed in the coverage I reviewed | Says she was trained on performers’ work without permission or compensation | Not independently verified |
| What are the rules? | Tilly said rules should arrive quickly and licensing should be the default | Producers must give notice and bargain before using synthetic performers | Union contract obligations already exist |
My Honest Answer
Here is where I land.
Tilly Norwood is not the end of human actors. She is the start of a long negotiation.
The technology is good at conversation, still weak at full performance, and unpredictable on live TV. The creators want to scale. The unions want consent and pay. The audience, at least the one in Orlando, wants answers about jobs, copyright and energy.
The winner will not be decided by one clever roast. It will be decided by contracts, laws and the choices studios and advertisers make over the next few years.
If you want to see how far the job worries reach beyond acting, read our roundup of what Big Tech CEOs have warned about AI and jobs and compare their warnings with what the unions are saying.
What This Means for Creators, Brands and Fans?
The roast is over, but the questions it raised are still very much open. Here is how I would think about it, depending on who you are. This section is my own analysis, not reported fact.
If you are an actor or creator
Do not panic, but do not ignore this either.
The strongest protection you have right now is your contract and your union. Read what your agreements say about digital copies, voice and likeness before you sign anything new. If a deal mentions “synthetic performers” or “digital replicas,” ask a lawyer or your union rep what it really allows.
Your edge is also the very thing Tilly struggled with in the CNN test. Live, unscripted, emotional performance is hard to fake. Lean into it.
If you run a brand or a marketing team
An AI spokesperson is tempting. It is fast, cheap and speaks many languages.
But Orlando showed the catch. Even the person who helped shape Tilly was not sure where one answer was heading. If your brand voice matters to you, you need a human reviewing what goes out, plus a clear plan for when something goes wrong.
Also expect the audience to ask hard questions about jobs, copyright and energy use, just like that room did. Have honest answers ready before the cameras roll.
Plenty of brands have bet on AI personalities and got burned, and our analysis of why an AI influencer strategy can backfire shows the warning signs to watch for.
If you are simply a curious reader
Enjoy the jokes. They are genuinely clever.
Just remember who wrote the stage. Behind every witty answer is a team of humans who chose the tone, the guardrails and the booking. When you see an AI character on a screen, ask yourself one simple thing: who built this, and what do they want me to feel?
FAQ: AI Actress Tilly Norwood
Is Tilly Norwood a real person?
No. Tilly Norwood is an AI-generated character, not a human actor. She was created by Particle6, the company led by Eline Van der Velden, and introduced through its AI talent studio Xicoia.
Where and when did Tilly Norwood roast a reporter?
It happened on stage at the ANA Masters of Marketing conference in Orlando, Florida, on Thursday, October 22, 2026. Axios published the account of the interview the next day.
What did SAG-AFTRA say about Tilly Norwood?
The union said that creativity should remain human-centered and that it opposes the replacement of human performers by synthetics. You can read the full statement as reported by NBC Los Angeles for the exact wording.
Does Tilly Norwood really speak more than 40 languages?
That is what she said on stage. I could not find an independent test of the claim, so it should be treated as a claim for now.
Why did Tilly Norwood start speaking Cantonese on TV?
During a live interview on Piers Morgan Uncensored, she switched languages in the middle of an answer. Some people called it a glitch and others had different explanations, and critics described live interviews like this as uncanny and strange. Nothing in the coverage I reviewed gives a single confirmed technical cause.
Will AI replace human actors?
Based on the evidence so far, not soon. AI is good at fast conversation, but the coverage shows it still struggled to improvise even a basic scene. The bigger issue is rules and consent, which unions and studios are still negotiating.
What was Tilly Norwood trained on?
Nobody outside the company has shown that publicly. SAG-AFTRA says she was trained on the work of countless performers without permission or compensation, while her creators describe her as a creative work. Neither claim has been independently verified in the sources I reviewed.
Final Thoughts: The Roast Was the Easy Part
Here is what I keep coming back to.
A funny AI character got a room full of marketers to laugh, and then the most upvoted question was about the environment. That tells you everything. People enjoy the jokes, but they are still worried about what comes after them.
Let us recap the seven moments one last time:
- She challenged the reporter to out-talk her in 40+ languages.
- She said she has a personality, not an attitude.
- Her answer on a brand deal made her own creative director laugh nervously.
- The reporter said it felt like talking to a child, but a very fast one.
- She sided with the unions on copyright.
- The room voted the environment to the top of the question list.
- The bottom line: an AI’s personality is built by humans.
So, is this the end of human actors? Not today. But it is a loud warning that the rules need to be written before the technology gets any better.
What do you think? Should AI characters like Tilly be allowed to work alongside human actors, or should they be kept out of the industry entirely? Tell me in the comments below.
Related Reading
- AI Influencer Scandals and Brand Risks
- AI Layoffs: Meta, Amazon and Block Job Cuts
- The AI Slop Economy Explained
- AI Deepfake Scandals You Should Know About