Workflow
AI Makes Entertainment Cheaper to Make. Forbes Contributor Asks Who Owns It and Who Gets Paid
· RenderBob team
Novika Ishar, writing in Forbes, argues that cheaper AI production does not by itself mean more profit. Ownership, brands and performer consent take more of the value.
Novika Ishar, a media and entertainment lawyer writing as a Forbes contributor, published an opinion piece on 10 November arguing that cheaper production does not by itself mean more profit. It is commentary, not reporting, and not legal advice. We read the Yahoo reprint.
The argument
- Cost is falling. Citing Reuters, the piece says production costs for AI-generated short dramas in China fell from about 5,000 yuan per minute to a few hundred yuan in the first half of 2026. A Reuters item we could read attributes that drop to state broadcaster CCTV. We have not read CCTV.
- Value moves to rights. As production gets cheaper, Ishar says ownership of IP, distribution relationships and recognisable brands may matter more.
- Copyright gaps. The piece says purely AI-generated material is not protected under US copyright law without sufficient human authorship, citing the US Copyright Office. Human contributions to an AI-assisted work may still qualify, which could complicate sequels, licensing and adaptations if some elements lack protection.
- Infringement and provenance. Producers could face claims if output includes protected characters or images, so the piece stresses checking where training and source data came from.
- Performers. Digital replicas and synthetic performers raise consent and pay questions, and permission for one project may not cover later productions or advertising. The piece cites California AB 2602 and AB 1836 and SAG-AFTRA's digital-replica terms. We have not opened the statutes.
- Contracts. Distributors may ask whether a production uses synthetic performers, third-party likenesses or tools with commercial-use limits, which could affect warranties, indemnities and insurance.
Ishar describes the Lionsgate–Runway custom-model deal as possibly signalling a trend. That is her speculation.
What a production team can keep on file
The article is general. Its points translate into a short file of records a producer could keep for each project.
- Record human authorship: who directed, edited and approved each shot.
- Record tool terms: which tools were used and what their commercial-use terms said on the date of use.
- Record sources: references, uploaded material and any likeness or voice permissions.
- Keep consent paperwork for any digital replica, with its scope and end date.
- Ask distributors what they want disclosed before delivery.
What to watch
Whether distributors begin asking for provenance records as a delivery condition, and whether the US Copyright Office or courts say more on how much human input is enough.
RenderBob 3 and this story
RenderBob 3 keeps references and prompt logs with the studio and gives generated clips their own timeline lane, which is the kind of per-shot record the rights questions in this piece depend on.
More from the blog
- Two Years, 1.2 Million Images, 100 Full-Film Passes: What Bona's Account of Sanxingdui Says About Long-Form Consistency
Bona Film Group has published numbers for how its all-AI feature was made. They are company figures, but they show where the effort in a feature-length AI film goes.
- What Claude Code Can and Can't Do in DaVinci Resolve: Five Tasks and What Went Wrong
A write-up of hands-on tests lists five editing tasks an AI coding assistant can drive in DaVinci Resolve, and records where it failed. Useful for rough work, with a person still finishing.