OpenAI has effectively built a social network where users can create digitized versions of their faces and voices, ready to be dropped into any prompt. That means anyone can place their avatar into whatever scenario they can imagine, from strolling through London in Autumn to starring in an anime. Or they can leave themselves out of it entirely and just turn stray thoughts into prompt-based videos. From there, it works like any other social network: users can edit, remix and respond to clips from friends, followers or whoever happens to show up in their feed. 

What once took creators of all sizes a mix of time, talent and trial-and-error now takes a few taps on Sora. Call it vibe creation — less about producing content and more about prompting it. And like vibe coding before it, it’s shaping a culture where creative output isn’t made so much as generated, one aesthetic impulse at a time.

Creators, predictably, are torn — intrigued, confused and a little terrified by what comes next. Deepfakes, digital replicas, blurred ownership, disappearing authorship — the lines between creativity, consent and control are collapsing in real time. To make sense of it all, some are already turning to peers who’ve already learned how to bend these new AI tools to their advantage before the tools start bending the definition of a creator itself.

The age of the vibe creator

In fairness, this reckoning started long before Sora. The relentless churn of AI breakthroughs over the past two years has forced creators to rethink what “making” even means. Sora just turned that quiet anxiety into something visible — and viral. How creators respond will determine whether they adapt to the moment or get consumed by it.

“There are three kinds of creators when it comes to AI: those who’ve mastered the tools, those who are learning fast, and those still operating business as usual,” said Jamie Guffreund, owner of consultancy Creative Vision.

Whoever they are, the clock is ticking. Maybe not as fast creators fear but certainly faster than they think. That’s been the through line in conversations with creator economy execs this week: Sora itself might not be the future — not yet, at least — but it’s a clear signal of where things are headed. Especially when you remember that Mark Zuckerberg has never met a social format he didn’t want to clone. 

And in that new world, the parasocial relationship that defined the creator economy’s most transformative phase will mutate again — this time into something far more uncanny. Imagine creators building followings through AI-generated videos that feature their followers as cameos, co-stars, love interests, or even as characters in stories they never approved. The creator becomes the architect of someone else’s fantasy — whether that someone is a fan, a brand or something much more manipulative. 

“If you think about the overall amount of time that any individual spends with creators, it’s going to be cut into by AI-generated content,” said Jim Louderback, editor and CEO of his “Inside The Creator Economy” newsletter. “That means there’ll be less watchtime hours to go around for creators, which means they will get disrupted at some level.”

That’s if — and it’s a big if — people actually want AI slop in their feeds to begin with. So far, the consensus leans no: people still crave the imperfection and personality that make content work on Instagram, YouTube and TikTok. Then again, history shows that nostalgia rarely wins out against novelty for long. As my colleague Tim Peterson explored here, once platforms find the engagement lever, the rest is just a matter of time. 

The bigger question for OpenAI — and anyone chasing the “social AI” — is what that lever even is. Posting used to be impulsive, now it’s strategic. Most people have quietly decided the costs of being visible online — the scrutiny, the judgement and the permanence — outweigh the upside. And that might be Sora’s real hurdle. 

The internet already runs on a handful of people willing to keep producing while everyone else scrolls. AI just makes that imbalance starker — easier to create, harder to care. And that imbalance hits creators hardest. Their job has always been to make people feel something, not just feed the algorithm. Because brands don’t hire creators to generate more content. They hire them to make it matter. 

Which is what makes the rise of vibe creators both fitting and a little tragic. As AI floods the feed with synthetic mods and manufactured moments, the creators who endure will be the ones who can still make those prompts — and the feelings behind them — feel human. 

“Perceptions of AI vary widely across the creator economy so I can’t speak on behalf of creators as a whole,” said Adrienne Lahens, co-founder and CEO of entertainment consulting firm Infinite Studios. “But what I am sure of is that we’re seeing the emergence of a new class of creators who use AI as their primary medium. AI is expanding who gets to be a creator.”

That comes with its own mix of risk and reward for marketers watching this unfold. Whether audiences want them or not, ads will be an unavoidable feature of this new kind of social media. That much is inevitable given how capital-intense and profit-light AI businesses are right now. Platforms like Sora — or the clone sure to follow — are shortcuts to gathering more user data and serving more precisely targeted ads. 

Creators, once again, are part of that product. Where they fit within it will depend as much on where the ads do. After all, the creator boom was, in part, backlash to how advertising had overrun the social web. The same pattern may repeat. Only this time, creators may be even more crucial to how brands show up.

A recent conversation with MediaMonks co-founder Wesley ter Haar explained why. He questioned whether CMOs would be interested in whatever ad format Meta, Sora et al can concoct over something more *sigh* “authentic”. It feels like the worst place to do a ‘normal’ ad, he told my colleague Sam Bradley. 

Which might be exactly why vibe creators will endure.

And another thing….

All of that analysis and we haven’t even touched on the most immediate fallout: the threat to creators’ intellectual property. In the days since it launched, the app’s been flooded with videos built from copyrighted material — faces, designs, footage — none of which the user had permission to use. Digiday’s platforms reporter Krystal Scanlon caught up with Steven Stein, partner in Greenberg Glusker’s Entertainment Litigation and Intellectual Property Litigation Groups, to unpack the implications.

This conversation has been lightly edited for clarity.

How are creators and brands responding to Sora? 

There are very few things that unite the city [LA], Hollywood and content creators, but taking content and not paying for it and reproducing that content is one of them. Sora 2 is one of the rare things, because unlike some of the other AI systems, Sora 2 reproduced copyrighted materials, including characters, and allowed people who are using Sora to do whatever they wanted with those characters. 

Are they exploring whether they’re going to be able to fight this really aggressively if the app has used their IP without consent? Or do you think we’re going to see them increasingly work to get in on the revenue-sharing arrangements? 

Rights holders are not going to just let this stand. The ways you can deal with this is by working with the AI companies or pursuing legal action against them. Sometimes, when you pursue legal action, that can lead to a deal being reached. But if you’re a content creator or a rights holder, are you going to license your content to AI systems? And if so, what’s the scope of that license? And it’s also possible that if the content is being used, you might want to be paid for those uses.

And I think a lot of AI companies, they’re rushing to the market, and many of them are asking for forgiveness and not permission. But ultimately, I think there’s a good chance you’re going to see legal claims. The potential liability is hard to fathom because every time there’s a use that infringes a copyright, it could lead to substantial damages, attorneys fees, etc. 

The legal framework for AI that is being worked out by the courts seems to go lighter on the input and focus on the output. Has the creator industry learned anything from the entertainment industry allowing Big Tech to dictate the rules of online content exploitation?

Creators, much like people in Hollywood, want to be paid for the work they’re doing. So, a lot of these issues implicate other issues. But tech companies that are in charge of, or own or operate AI systems, or invest in them, create an existential threat to content creators. 

In the three AI cases in the U.S. that resulted in trial court decisions, the key issue in one of them was, when you’re using content to train AI, is that transformative enough to be protected by the fair use doctrine? And it was, it’s a fact-specific inquiry. But one court said it’s spectacularly transformative, because you’re taking books, and aspects of those books are being used to train AI about all different things. And AI isn’t necessarily reproducing those books verbatim, provided there are sufficient protective measures in place. 

But none of those cases really dealt with outputs where the content was replicated exactly by the AI systems. There’s The New York Times versus OpenAI case, where The New York Times was able to point to examples where articles were replicated by its outputs, and OpenAI says, well, you have really had to contort the system to get there. That’s going back and forth. 

Typically, the reason content creators can take things and comment on videos is because they’re protected by the fair use doctrine that they’re transforming the content by maybe only using a small portion of it or commenting.

This is why copyright exists. This is a fundamental reason why you get a copyright: we want to incentivize people to create. We want to incentivize people to be able to monetize what they create. That’s one of the reasons people create content. And so if you’re making it less valuable to create content, because AI can just do what it wants with the content you’ve created, without paying you money to use that content that could disincentivize people to create content. And so I think that’s a fundamental dynamic that Sora 2 exposed.

So far, the lawsuits have been fairly nebulous about whether training an AI model is protected under fair use, but they seem to be pointing to the notion that infringing output is a bit of a different story. From a legal perspective, does Sora seem a more straightforward target?

Yes, because none of those cases dealt with outputs. In the Meta case, the judge credited Meta with having guardrails in place to ensure that content that was being inputted wasn’t being outputted. Copyrightable, copyrighted content. 

What we’ve covered

Vibe marketing was a joke – until OpenAI got involved

At OpenAI’s first developer conference in two years, CEO Sam Altman rolled out a slate of updates but one message was clear: the company wants ChatGPT to be the operating system for everything.

Why advertisers are quietly returning to news-driven media channels

Brand safety was more about brand suitability, but under the new President Trump administration, marketers are having to grapple with a fast-moving threshold.

Google’s ad tech antitrust remedy phase explained — and where the judge might land

While Digiday wasn’t in the courtroom as the Department of Justice’s antitrust case wrapped on Oct. 3, the team has been trading notes with people who were to get the lowdown on what really matters.

What we’re reading

Advertisers push big tech to adopt standards for transparency in ad sales

According to the WSJ, some of the industry’s biggest players have come together to propose new standards for transparency in digital auctions that are increasingly dominating ad sales. Why does this matter? Auctions don’t always go to the highest bidder, and the price that’s ultimately paid isn’t always what the winner actually bid.

Goldman Sachs, JPMorgan flag rising debt levels at AI companies

Unsurprisingly running an AI company isn’t cheap, with both Goldman Sachs and JPMorgan flagging the record debt already racked up this year, according to The Information. It’s not a problem — yet, but something the industry needs to keep an eye on. 

What the arrival of AI video generators like Sora means for us

While the likes of Sora democratize the ability for anyone to create content, the problem is, it’s already enabled some users to generate fake crimes, celebrity brawls and deepfakes of the dead, according to The New York Times. Ultimately, in less than a week, Sora has already managed to rewrite digital trust.