In a legal fight that could set the price of AI training data, Hollywood’s creative establishment is challenging one of the technology sector’s core arguments: that licensing copyrighted work at scale is “impossible.” A coalition led by Cate Blanchett’s Human Consent Foundation filed an amicus brief Friday in the Southern District of New York in The New York Times’ copyright suit against OpenAI and Microsoft.
Who joined the filing
The 31-page document was co-signed by CAA, the Music Artists Coalition, the National Association of Voice Actors and the Creators Coalition on Artificial Intelligence. An appendix carries support from Anne Hathaway, Bette Midler, Emma Thompson, Helen Mirren, Dan Levy, Ron Howard, Madonna and Olivia Wilde.
The core argument
The defendants have argued that clearing creative works for AI training is cost-prohibitive and that there is no settled method to value training data. The brief pushes back by pointing to an infrastructure that already exists: collective licensing. For more than a century, organisations such as BMI and the International Confederation of Societies of Authors and Composers have administered rights across large repertoires using blanket licences and metadata-driven systems, the filing says.
Blanchett, who has worked across film and theatre for over three decades, says the principle has been consistent: permission, contract and conversation before someone uses her face, voice or image. “Consent isn’t a barrier to creativity, it is the foundation of it,” she said.
Why this matters in the NYT case
The Times sued OpenAI and Microsoft in December 2023, accusing the companies of using its journalism to build ChatGPT and Copilot without permission or payment, and of creating products that compete directly with its reporting. The new brief follows a separate filing from SAG-AFTRA, the Writers Guild of America and the American Federation of Musicians that urged the court to reject a categorical fair-use exemption for AI training. Together, the filings signal that Hollywood labour and talent agencies are aligning around licensing as the baseline.
- The brief’s central claim: collective licensing already solves the “dispersed owner” problem through centralised clearinghouses.
- The practical effect: if courts see licensing as workable, “cost-prohibitive” becomes a weaker defence.
- The star-powered appendix adds public and industry weight to a case otherwise framed as a publisher-vs-platform dispute.
What to watch
For rights holders and media planners, the filing is a reminder that AI training data is becoming a negotiated asset rather than an assumed input. Companies with clean chain-of-title records and centralised rights data will be better positioned to license catalogues at scale, whether directly or through a collective. The risk of waiting is that courts and platforms set the default before individual creators and publishers do. That makes the outcome not just a legal question but a valuation benchmark for every catalogue owner.
The case now carries two competing framings: AI firms describe training-data licensing as broken and impractical; the creative coalition describes it as a century-old operational process. The answer the court accepts will influence how streaming, news, music and film libraries are priced for machine learning.
Source: TheWrap
