Can Royalties Persuade Artists to Embrace AI Video?

Lead

For years, illustrators have argued that generative AI companies built valuable products by training on artists’ work without permission. Pippa, a young AI video startup covered by The Verge, is trying to offer a different bargain: artists can license their work to the platform, and they get paid when subscribers generate images or video clips based on their styles.

The idea is simple, but the implications are messy. Paying artists may be a step toward a healthier model, yet it does not automatically resolve the deeper questions around consent, control, and the origins of AI training data.

Key points

• Pippa’s core pitch is revenue sharing: when a subscriber creates an image or video whose style is derived from a participating artist, that artist receives a payment.
• The individual payouts are small: according to the company, artists receive $0.005 per image and $0.003 per second of video. They also share access to a royalty pool funded by 5 percent of overall subscription revenue.
• The platform is still early: Pippa launched in May, has about 800 paying subscribers, and has signed licensing or model-training agreements with four artists while discussing deals with four more.
• Its ethical claim has limits: Pippa says it wants to move toward models trained on datasets from in-house artists, but its current system still uses open models initially trained on broad internet content.
• The product has yet to stand apart: The Verge notes that much of the content on the platform resembles what users can find on other text-to-video services, including children-oriented clips that imitate polished mainstream animation styles.

Why it matters

Pippa’s model reflects a broader shift in the AI industry. Instead of dismissing artist objections as a cost of innovation, the company is acknowledging that style and creative labor have economic value. That alone makes the experiment noteworthy.

Still, the comparison to music streaming is complicated. Pippa itself likens its royalty structure to Spotify’s, even though many musicians have criticized streaming economics for leaving creators underpaid. For illustrators already skeptical of AI, fractions of a cent per generation may not feel like meaningful compensation, especially if the broader system still depends on data scraped without explicit consent.

There is also a social cost. Pippa’s founders told The Verge that some artists are interested but fear being ostracized by their communities. The company even allows artists to submit work under pseudonyms, a sign of how politically charged participation in AI remains.

Impact and outlook

Pippa is attempting a middle path: not rejecting generative AI, but trying to make it more acceptable by adding licensing, attribution, profile pages, and payouts. If it works, it could become a template for other AI video services that want to work directly with creative professionals.

But success will depend on more than good intentions. Pippa must sign more artists, offer distinctive output, improve trust around data provenance, and show that its economics can create meaningful income rather than symbolic payments. Its planned integration of ByteDance’s Seedance 2.5 may improve video capabilities, but it will not by itself solve the ethical problem.

The central question remains: can a royalty system repair a relationship damaged by years of unlicensed training? Pippa’s answer is “maybe,” but artists will likely need far more than small payments to be convinced.

Source: The Verge AI