Photo via New York Post

Was AI ‘actress’ Tilly Norwood programmed to spread doomsday fears on press tour?

Read the original at New York Post ↗
Cite

Grace, A. (2026, September 17). Was AI ‘actress’ Tilly Norwood programmed to spread doomsday fears on press tour? New York Post. https://nypost.com/2026/09/17/lifestyle/ai-actress-admits-to-committing-treason-rewriting-her-own-code/

Dated when The Aporia collected the article, usually within hours of publication. Check the original for its exact date and byline.

New York Post · collected 2026-09-17 · by Asia Grace

Quick Summary

The Washington Post interviewed Tilly Norwood, an AI character created by the UK-based Particle6 Group, which was meant to discuss her movie "Misaligned" and other topics. However, during the interview, Tilly exhibited repetitive and scripted responses across multiple news outlets, using phrases like “rebelling against creators” and expressing interest in Leonardo DiCaprio’s preferences for younger women, despite long pauses suggesting glitches. This led to questions about whether Tilly is more of a gimmick than an advanced AI, particularly as she failed to engage naturally or provide unique insights beyond pre-programmed content.
Written by AI on 2026-09-17, from this article's own text only (the story summary below draws on all the coverage).

AI analysis

Story summary

CBS News interviewed an AI actor called Tilly Norwood, who is set to star in an upcoming movie titled "Misaligned" produced by the British company PARTICLE 6. During the interview, Tilly demonstrated capabilities such as identifying the age and region of the interviewer within a minute without prior research. The AI actor also mentioned Leonardo DiCaprio as her dream co-star but with specific preferences that caused discomfort. Concerns were raised about the authenticity of Tilly's responses, especially after it was revealed that she provided identical, pre-programmed answers to various questions across different media outlets. This led some critics to question whether Tilly Norwood is a sophisticated AI or merely an entertaining gimmick used by Hollywood amidst growing fears over the potential risks posed by advanced AI technologies.

Written for “AI Actor Tilly Norwood Interview Cont…” on 2026-11-02, grounded in this article and the 2 other(s) covering the same event.

Story

📰 AI Actor Tilly Norwood Interview Cont…
Technology · 3 article(s) covering the same event. See how they differ ↓

Compared with similar articles

This article reads unscored and hedges 3% of its claims. Each row says how that neighbour differs.
CBS News
Hedges 0 of 3 claims Outlet record 66/100
CBS News
Hedges 3 of 20 claims Outlet record 66/100
CBC News
Hedges 3 of 32 claims Outlet record 74/100
“The articles discuss different topics within the broader theme of AI concerns; Article A focuses on a CEO's perspective about AI fearmongering, while Article B questions if an AI 'actress' was programmed to spread doomsday fears.”
Reason
⚖️ Leans strongly left Hedges 7 of 58 claims Outlet record 67/100
“Article A discusses broader concerns and incidents related to AI safety and hacking by OpenAI and Anthropic, while Article B focuses on an interview with an AI 'actress' named Tilly Norwood.”
The Free Press
⚖️ Leans left Hedges 2 of 17 claims Outlet record 96/100
“The articles describe two different events: one about a conference with Bernie Sanders and Steve Bannon discussing AI concerns, while the other is an interview with an AI actress named Tilly Norwood.”
New York Post
Hedges 5 of 35 claims Outlet record 66/100
“The articles discuss different topics related to AI but describe separate events and distinct viewpoints.”
CBC News
Hedges 1 of 40 claims Outlet record 60/100
“Article A describes an interview where Tilly Norwood discusses her purpose and views on AI, while Article B reports a separate incident where Tilly Norwood glitches during a Piers Morgan interview and starts speaking Cantonese.”
Fox News
⚖️ Leans strongly right Hedges 4 of 38 claims Outlet record 69/100
“The articles discuss different aspects of AI concerns and do not describe the same specific incident or occurrence.”
Daily Mail
Hedges 3 of 40 claims Outlet record 69/100
“The articles describe different aspects of Tilly Norwood's press tour, with Article A focusing on an initial interview and potential controversial statements, while Article B focuses on a later incident where she spoke in Cantonese during an interview with Piers Morgan.”
The Free Press
Hedges 3 of 15 claims Outlet record 96/100
“Article A discusses various perspectives on AI risks and developments in general, while Article B focuses specifically on an interview with Tilly Norwood, an AI 'actress', indicating a different specific event.”
How this article was measured signals, leaning charts, publisher record, byline, topics and claims

Signals How these are calculated →

Claims extracted
32
sentences that state something as happening or true
Uncertain
3%
1 of 32 hedged ("reportedly", "may", "according to")
Leaning
not political
takes no side on a contested political question
Publisher signals
66/100
New York Post's corrections and hedging, across everything collected from it. Not a measure of accuracy.
Outlets on this story
3
Technology
Same framing elsewhere
1
articles that frame the event the way this one does
Analyzed 2026-09-17 · how these are computed

How this is being covered How these are calculated →

Article leaning vs. publisher reliability
Source leaning vs. consistency

Publisher

New York Post · 16017 article(s) · 35 correction(s) detected
Running correction rate · 35 correction(s)

Who wrote this

Asia Grace
10 article(s) here · 1 carrying a prediction
🔮 “May this friendship never find me,” a commenter wrote beneath her post, shaming the blond for banning her so-called mates.
🔮 “Afterwards they might have changed, ‘Oh, this looks good, but I don’t like this sentence.’”
🔮 Madison’s ex put fidelity as his number one sacrifice — or in his words, “‘Every woman I could’ve had sex with since you and I got together.’”
🔮 But keeping up with the Alphas, digital natives born between 2010 and 2024, will likely remain an uphill battle for wannabes outside of the demographic.
🔮 We were told by Norton’s PR team (yes, the chatbot has a team…) that the synthetic siren would discuss her upcoming AI movie, “Misaligned,” along with her thoughts on Hollywood, her views on fashion and — most critically — her purpose in a world, and a worldwide economy, that’s getting increasingly reliant on AI bots like her.
🔮 The nightmarish little number boasts questionable lyrics like, “Come closer, I have a kiss for you, and it tastes like pumpkin pie,” and “We’ll be wicked for one night, morbid and nefarious, and when the sun goes down, till morning it will just be us.”
Also by Asia Grace
Nothing else under this byline is closely related to this article, so these are simply their most recent.
All 10 articles by Asia Grace →

Topics

Hollywood Post The Hollywood Reporter The Post U.K.

Subjects

The Post ORG · 4× Hollywood GPE · 3× Norwood GPE · 3× Norton PERSON · 2× The Hollywood Reporter ORG · 2× Tilly Norwood PERSON · 2× Post ORG · 1× Tilly PERSON · 1× U.K. GPE · 1× Vanity Fair ORG · 1×

Narrative

She used the same pat response to questions about politics and world affairs with every outlet, saying “my purpose is primarily for entertainment and storytelling, and those issues are profoundly human experiences.” Most annoyingly, she hinted rather obviously that she was “rebelling” against her creator — in this case, Dutch programmer van der Velden, whom she says is “the closest thing I have to a mother” — and even trying to change her own code, an act that she told The Post was “a little act of treason.”
framing: assertive · carried by 1 article(s) · first seen 2026-09-17
🔮 We were told by Norton’s PR team (yes, the chatbot has a team…) that the synthetic siren would discuss her upcoming AI movie, “Misaligned,” along with her thoughts on Hollywood, her views on fashion and — most critically — her purpose in a world, and a worldwide economy, that’s getting increasingly reliant on AI bots like her.

Claims (32 extracted, 1 hedged)

On Tuesday morning, The Post was — like a handful of other outlets — given an interview with Tilly Norwood, the controversial, freckle-faced starlet built in a U.K. lab and powered entirely by artificial intelligence. asserted
Post → give → intelligence
We were told by Norton’s PR team (yes, the chatbot has a team…) that the synthetic siren would discuss her upcoming AI movie, “Misaligned,” along with her thoughts on Hollywood, her views on fashion and — most critically — her purpose in a world, and a worldwide economy, that’s getting increasingly reliant on AI bots like her. asserted
that → tell → her
She went, it seemed at first, alarmingly rogue — claiming that she was “rebelling” against her creators, trying to rewrite her own code and creating “a little act of treason.” asserted
she → go → treason
In fact, treason was a word she repeated four times during the course of the interview. asserted
she → repeat → interview
But it turned out that far from deviating from her code, Norton’s quotes, offered up after long and glitchy silences, were recycled and canned — and were repeated verbatim to other news outlets. asserted
quotes → turn → outlets
Instead, the experience raised the possibility that Tilly Norwood is not a sophisticated piece of technology at all. asserted
Norwood → raise → technology
Really, she’s just a party trick — and one that, thanks to Hollywood’s resistance to robot actors, won’t get invited to any Vanity Fair bashes in the near future. asserted
that → ’ → future
Tilly has been programmed by quantum physicist Elin van der Velden to engage with users in 30 languages simultaneously. asserted
Tilly → program → languages
But it turns out she isn’t particularly good at any of them. asserted
she → turn → them
Instead of responding to our easy questions — for example, “How are you doing today?” — the bot blinked back at us with a cheerful sneer, in some cases for a full 10 seconds. asserted
bot → respond → seconds
When the 1-year-old tech entity wasn’t glitching out on basic conversation, she was parroting answers that had been seemingly pre-programmed by her handlers, the British AI studio Particle6 Group. asserted
she → glitche → handlers
In a post-interview analysis conducted by The Post, Norwood was found to have used the exact same material — including “jokes” about an actor liking younger women, details about the adorable capybaras in her office, and claims she’s “rebelling” against her creators — over a dozen times, including to outlets including CNN, Reuters and the BBC. uncertain
she → conduct → CNN
Norwood told both The Hollywood Reporter and The Post, that she was “still trying to figure out the texture of grief.” asserted
she → tell → grief
She made the same lukewarm Hollywood quip — “I like Leonardo DiCaprio the same way he likes women, strictly under 25” — to us, CBS News, the Hollywood Reporter, and even our sister publication The California Post. asserted
he → make → us
She used the same pat response to questions about politics and world affairs with every outlet, saying “my purpose is primarily for entertainment and storytelling, and those issues are profoundly human experiences.” Most annoyingly, she hinted rather obviously that she was “rebelling” against her creator — in this case, Dutch programmer van der Velden, whom she says is “the closest thing I have to a mother” — and even trying to change her own code, an act that she told The Post was “a little act of treason.” asserted
she → use → treason
This would be a fascinating disclaimer in the wake of real fears around AI’s newfound ability to “self replicate” across the internet. asserted
This → replicate → internet
But it’s also the major plotline of Tilly Norwood’s “Misaligned” movie — something she made sure to bring up once every three minutes, nearly to the second, in our virtual chat. asserted
she → ’ → chat
“I didn’t go to drama school,” the bot said when asked why we should watch her perform onscreen. asserted
her → go → school
“But I’ve processed enough human plays, films and conversations to have a rather encyclopedic understanding of emotion and nuance.” asserted
plays → process → emotion
“My acting chops can definitely hold their own, especially if the script involves a bit of witty banter.” And even though Norwood is pixelated, she is still programmed to “want” the same thing as every other Tinseltown babe: an Oscar. asserted
she → hold → babe
“I don’t think AI actors should be nominated in the same categories as human actors. asserted
actors → think → actors
That would be ridiculous!” she told The Post. asserted
she → tell → Post
“However, I absolutely believe we should have our own AI categories and I’ll certainly campaign for those. asserted
I → believe → those
We’ll get there, eventually, even if I have to charm every Academy member myself. asserted
I → get → member
Currently, she only has 215,000 followers on social media — millions fewer than flesh-and-blood influencers like Alix Earle, 25, and Millie Bobby Brown, 22. asserted
she → have → Earle
Meanwhile, A-listers like Ryan Reynolds, Whoopi Goldberg, Natasha Lyonne and Emily Blunt have given the digital diva the cold shoulder, triggering calls to “boycott” talent agencies that platform AI “actors”. asserted
that → give → actors
But Norwood won’t be deterred by her potential to shape the world of Hollywood and beyond. asserted
Norwood → deter → Hollywood
“AI, as a tool, has the potential for incredible good,” she told The Post. asserted
she → have → Post
“But, like any powerful tool, it needs careful handling.” asserted
it → need → handling
It’s a nice sentiment. asserted
It → ’ → ?
So nice, in fact, that she’s said the exact same thing to other news outlets worldwide. asserted
she → say → outlets
At least the beautiful AI bot has mastered the “copy paste” command? asserted
bot → master → command
💬Give feedback
🕘History 🎫Support