Pippa bets artists will embrace AI if the royalties are right

The News

Pippa, a text-to-video startup that opened its doors in May, is betting on an idea most of its rivals have dodged: actually paying the illustrators whose styles power its models. Cofounders Hogan Shrum and Sean Wright built a revenue-share system that pays an artist each time a subscriber generates an image or clip derived from that artist's work, The Verge reports.

The individual sums are small. Pippa pays $0.005 for every image generated and $0.003 per second of video, topped up by a share of a 5 percent royalty pool funded from subscription income. Plans cost between $14.99 and $99.99 a month. The firm compares its model to Spotify, a curious benchmark given how bitterly musicians have complained about streaming cheques.

The operation is still tiny. Pippa counts roughly 800 paying subscribers and has signed licensing and model-training agreements with only four artists, with four more in discussion. Each illustrator is vetted to prove they own the work submitted, and may publish under a pseudonym to dodge accusations of "crossing the picket line". The company intends to plug in ByteDance's Seedance 2.5 model, which can produce up to 30 seconds of footage in one go.

Why It Matters

The launch arrives amid open legal warfare between creators and the firms scraping their output. Wright frames Pippa as a break from what he calls the industry's bloody history, reaching for a familiar analogy: music piracy on Napster ended not through litigation alone but once Apple made buying a track for 99 cents easier than stealing it. The wager: a workable payment rail, not a courtroom, is what finally normalises AI content licensing.

The catch is that Pippa has not solved the problem it markets against. Its current tools still lean on open models that were, in the company's own words, initially trained on the broad sweep of internet content, much of it taken without consent. Building a genuinely clean dataset is expensive; Ben Affleck's InterPositive has gone that route with proprietary data, but the cost pushes such firms towards narrow tools, not mass-market ones. Pippa's candour about the gap underlines how far "ethical AI" still is from settled.

Indian Angle

For India, the timing is pointed. The Delhi High Court is already hearing news agency ANI's suit against OpenAI over the alleged use of its content for training, and Indian music labels grouped under the Indian Music Industry body have pushed to join that fight. A working micro-royalty rail like Pippa's could become Exhibit A for rights holders arguing that paying creators is technically feasible, not fantasy.

The economics, though, would sting. At $0.005 an image, an Indian illustrator earns well under half a rupee per generation, a rate that only adds up at volumes few artists will ever reach. India holds one of the world's deepest benches of freelance illustrators, animators and VFX talent, much already working cheaply for studios abroad. A Spotify-style pool risks importing the very complaint Indian musicians make about streaming: the platform scales, the creator does not.

There is also a policy opening. MeitY and the courts are still shaping how India treats training data and creator consent. A licensing template built abroad, however imperfect, hands Indian regulators and homegrown startups a concrete reference point rather than an abstract debate.

FAQ

How much does Pippa pay artists?

$0.005 per generated image and $0.003 per second of video, plus a share of a 5 percent royalty pool funded from subscription revenue. Payments rise with usage, so higher-volume styles earn more.

How is this different from fully licensed rivals?

Firms such as Ben Affleck's InterPositive build proprietary datasets from scratch, avoiding scraped content. Pippa still uses open models trained partly on internet-scraped art while it onboards in-house artists, making it a hybrid rather than a clean-slate approach.

What does it mean for Indian creators?

Potentially a template for consent-based payment, but at rupee-level rates that reward scale over individual craft. India's large freelance base could gain a new income line or face another low-margin platform.

Is there an Indian legal angle?

Yes. The ANI versus OpenAI case in the Delhi High Court, plus interventions sought by Indian music labels, make training-data consent a live question for Indian courts and regulators.

Where can I read the original report?

The Verge published the original reporting by Charles Pulliam-Moore, linked below.

This story was reported by The Verge. Read the full original coverage at The Verge.