Equity Threatens Mass Action Over Unauthorised AI Use of Actors’ Images

TL;DR:

  • Equity threatening mass direct action over unauthorised AI use of members’ likenesses
  • Union plans coordinated data requests to force tech companies to disclose AI data usage
  • Growing complaints about copyright infringement and voice replica misuse

The performing arts union Equity has threatened mass direct action against tech and entertainment companies using its members’ likenesses, images, and voices in AI content without permission. General secretary Paul W Fleming announced plans to coordinate data requests en masse to force companies to disclose whether they used members’ data in AI-generated material without consent.

Context and Background

The warning comes as growing numbers of Equity’s 50,000 members have made complaints about infringements of their copyright and misuse of their personal data in AI material. Most complaints concern AI-generated voice replicas, though the union recently confirmed support for Scottish actor Briony Monroe, who believes her image was used to create the AI “actor” Tilly Norwood.

Fleming said the union has already helped members make subject access requests to producers and tech companies, which failed to provide satisfactory explanations about data sources. Under data protection law, individuals have the right to ask for all information an organisation holds about them, with organisations having one month to respond.

Looking Forward

Fleming said the tactic would create significant “hassle” for firms unwilling to negotiate, pushing tech companies and producers to strike agreements on performers’ rights. “AI companies need to know that we will be putting in these subject access requests en masse,” he said. “They have a statutory obligation to respond.”

Equity UK has been in negotiations with the UK production trade body Pact about AI, copyright, and data protection for over a year. Fleming noted that privately, industry leaders admit it is “impossible to ethically use AI because it has been scraped and trained on data where the provenance is at best unclear.”

Source Attribution:

Share this article