Technology news and discussion on TalkFuse.
What Happens When Documentary Makers Sell Their Footage to AI Aggregators?
Documentary specialists at IDFA discussed the risks and protections involved when filmmakers license archival footage to AI companies for training. As AI use becomes more common, the choices can affect both immediate income and filmmakers’ control over how their work and the rights within it are used.
Rights Stuff chief executive Wendy Bernfield said some AI aggregators include contractual protections, though they may pay less than services with greater rights risks. Archive producer Elizabeth Klinck said filmmakers approach her about AI enquiries almost weekly, and described a Canadian filmmaker who needed several months to change a contract’s five-day opt-out deadline. Klinck advised filmmakers to plan for future disputes, as AI use may complicate copyright clearances and errors-and-omissions insurance.
In short
- Documentary makers are being approached to license archive footage for AI training.
- Aggregator contracts vary in rights protections and fees.
- Short opt-out deadlines can leave filmmakers with little time to act.
Steelman: both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
Licensing footage to AI companies can give filmmakers and archives a new source of income and put existing material to productive use as demand grows. With clear contractual safeguards, including meaningful limits and opt-out rights, a deal may offer a reasonable balance between earning revenue and retaining control.
The case against
AI training deals can grant broad or difficult-to-reverse uses of footage, potentially reducing filmmakers’ control over their work and rights held by others. Contracts may set tight opt-out deadlines, and uncertain future uses could complicate copyright clearances or insurance, so filmmakers may reasonably prefer to withhold footage or accept only more restrictive terms.
Explained simply
Documentary films often use archival footage, such as old news reports or home movies, which may carry separate rights owned by people or organisations. Filmmakers can license this material to AI companies that gather footage to train their systems, in return for payment.
The terms of a licence can affect how footage may be used, whether a filmmaker can withdraw it and who is responsible if rights are challenged. These choices matter because a film may include many rights holders, and unclear permissions could create problems for later uses or insurance.
AI Entertainment Film Technology