1. Headline & intro
Tilly Norwood’s new AI power ballad is terrible — but that’s precisely why it matters. The synthetic “actor”, created by production company Particle6, has released a music video so awkward that TechCrunch’s reviewer called it the worst song they’d ever heard. It’s easy to laugh and move on. We shouldn’t. Tilly’s song is a neat compression of everything uneasy about generative entertainment: scraped labor, confused narratives about “AI rights”, and an industry that seems more interested in automating creativity than nurturing it. In this piece, we’ll unpack what this release reveals about the future of actors, musicians and audiences.
2. The news in brief
According to TechCrunch, UK-based production company Particle6, which introduced the AI-generated character Tilly Norwood in 2025, has now pushed the character into music. The company released a song and video titled “Take the Lead”, created by a team of around 18 people, including designers, prompt engineers and editors.
The track is sung from Tilly’s perspective as a misunderstood AI entity, complaining that critics dismiss her because she isn’t human and insisting on her own supposedly “human” qualities. The lyrics urge “AI actors” to seize their destiny and embrace a new evolutionary stage for performance. The video shows the digital character striding through a data centre and performing to a fictional stadium crowd.
TechCrunch’s reviewer harshly criticises the song as derivative and emotionally hollow, comparing it unfavourably with earlier AI-linked music projects like Xania Monet. The piece also recalls Hollywood union SAG‑AFTRA’s prior statement condemning Tilly Norwood as a product of training on actors’ work without consent or compensation.
3. Why this matters
Particle6 probably intended “Take the Lead” as a proof-of-concept: look, an AI actor who can sing, perform and anchor an entire music video. As a technical demo, it’s mildly impressive. As a cultural signal, it’s a flashing red light.
First, the song accidentally exposes how disconnected some AI evangelists are from actual audiences. The central premise — an AI lamenting that humans don’t recognise its humanity — is something no listener can authentically experience. Emotional resonance usually works because we can map it to our own lives. Here, that bridge simply doesn’t exist. The result is not just cringe; it’s a fundamental failure of storytelling.
Second, the project underscores who really benefits from AI “actors”. It’s not viewers, who already struggle to sift through endless mediocre content. It’s not human performers, whose work underpins these models but who are fighting just to be asked for permission. The winners are production firms that can build infinitely reusable IP with no residuals, no unions and no ageing.
Third, the rhetoric around the song is quietly dystopian. A synthetic persona, built on uncredited human labour, is positioned as an oppressed underdog rallying fellow AI to “take the stage”. It inverts the power dynamic: the technology backed by capital is framed as the victim, while the actual humans whose livelihoods are at stake are cast as the narrow-minded establishment.
Finally, Tilly Norwood illustrates the shift from AI as a tool for artists to AI as a replacement for artists. That’s the line many in creative industries hoped we wouldn’t cross. This song doesn’t just tiptoe over it — it dances on it with a key change.
4. The bigger picture
Tilly is not appearing in a vacuum. She follows a wave of AI-linked performers and virtual personas: from Hatsune Miku’s vocaloid concerts to Lil Miquela’s influencer campaigns and, more recently, AI-assisted artists like Xania Monet whose songs have charted.
The difference is where the human stays in the loop. Earlier virtual idols were ultimately vehicles for identifiable human songwriters, producers and voice actors. Even when heavily processed, their authorship was clear and often celebrated. With Tilly Norwood, the selling point is precisely the opposite: this is a character, a voice and a performance that can be generated and iterated without a single named performer in front.
That trajectory aligns with a broader industry trend: the industrialisation of content. Streaming platforms and social networks reward volume and retention, not depth. Generative AI is the perfect engine for filling every niche with “good enough” media — background music, filler TV, in-game chatter, algorithmic pop.
Historically, every new media technology has sparked moral panic — the synthesiser, drum machines, even sampling. But those tools expanded what human artists could do; they didn’t erase the notion that someone was behind the work. The push for AI-native actors and musicians threatens to normalise something different: creative products whose lineage is an opaque model trained on billions of unconsented contributions.
And this isn’t just about culture. It’s about bargaining power. The more studios can point to working AI stand‑ins, the easier it becomes to pressure human actors and musicians in negotiations. The fact that SAG‑AFTRA singled out Tilly Norwood last year is telling: unions see these characters not as fun curiosities but as bargaining chips deployed at the table.
5. The European angle
For European audiences and creators, Tilly Norwood lands in a much denser regulatory and cultural landscape than in the U.S.
The EU AI Act — now entering implementation — contains specific transparency obligations for AI-generated content and stricter rules around deepfakes. A fully synthetic “actor” fronting songs and films will almost certainly have to be clearly labelled as such, especially when distributed on major EU platforms. That undermines one potential advantage for producers: the ability to blur the line between human and machine.
Copyright and neighbouring rights add another layer. European collecting societies and organisations for performers’ rights (such as GEMA, SACEM, or PPL’s European counterparts) are already mobilising against unlicensed scraping of music and performances. If a character like Tilly is trained on European actors’ faces, voices or performances, rights-holders here are far better placed to challenge that in court than their peers in many other regions.
Culturally, Europe invests heavily in human-centred arts, from national theatres and orchestras to publicly funded film institutes. It is hard to imagine, say, ARTE or the Berlinale enthusiastically programming AI-fronted musical projects that lean on uncredited labour. European commissioners regularly speak of “cultural diversity” and “European authorship” as strategic assets; synthetic performers created in London or Los Angeles but built on global training data land awkwardly in that narrative.
For European studios, though, there is a temptation: AI actors as a way to stretch tight budgets and bypass language barriers with infinitely localised, lip‑synced faces. The question is whether EU rules — and public backlash — will make that path more trouble than it’s worth.
6. Looking ahead
What happens next depends less on the technology — which will rapidly improve — and more on three forces: regulation, unions and audience taste.
Regulators in Brussels, London and national capitals are already looking at how to operationalise AI Act provisions, platform transparency rules and copyright enforcement. Expect guidance, and possibly test cases, around synthetic performers within the next two to three years. A case involving a recognisable European actor’s likeness turning up in an AI character could move very fast.
Unions will push to lock in protections in their next bargaining rounds: explicit consent requirements for digital replicas, minimum compensation for training use, and restrictions on replacing human background actors and voice artists with AI. The Hollywood strikes of 2023–24 were an opening skirmish; characters like Tilly are ammunition for the next ones.
The wild card is audiences. If synthetic performers mostly generate mockery and indifference — as “Take the Lead” has so far — investors may quietly retreat, or reposition AI as back‑office tooling rather than the star. But if someone manages to build a compelling AI‑fronted act that teens genuinely obsess over, the economic pressure to prioritise machine‑based performers will surge.
In the meantime, creators face a strategic decision: ignore AI and risk being blindsided, or embrace it as a tool while drawing clear ethical red lines. The more artists visibly experiment with AI on their own terms, the harder it becomes for studios to argue that only fully synthetic, rights‑free characters can unlock innovation.
7. The bottom line
Tilly Norwood’s awful song is a punchline today, but it sketches a serious future: one where studios use synthetic performers built on uncredited labour to flood our feeds with frictionless, emotionally empty content. That future isn’t inevitable; it will be shaped by laws, contracts and what we choose to watch and stream. The real question is not whether AI can act or sing, but whether we are prepared to defend the value of human performance when the synthetic alternative becomes cheap, ubiquitous and — eventually — much better than Tilly.
