Some links on this page are affiliate links: if you buy through them we may earn a commission, at no extra cost to you.
On January 22, 2026, the campaign Stealing Isn’t Innovation publicized an open appeal signed by more than 700 creatives, including Scarlett Johansson, Cate Blanchett, R.E.M., Vince Gilligan, Kristen Bell, Common and Cyndi Lauper. The signatories object to creative work being used to develop AI systems without authorization and call for licensing and partnerships—not an outright ban on AI. The campaign’s phrase “theft at a grand scale” is its characterization of the alleged conduct, not a legal finding.
What the campaign is asking for
Stealing Isn’t Innovation argues that AI companies should not treat creative work as free raw material for commercial systems. Its proposed alternative is authorized access through licensing agreements and partnerships with creators or rights holders, with compensation where appropriate. The campaign says AI can continue to develop while respecting creators’ rights.
Its objections reach beyond training data. The campaign also raises concerns about the economic value captured by AI companies, the unauthorized use of voices and likenesses, deepfakes and other forms of misinformation, and mass-produced material it calls “AI slop.” These concerns overlap, but they are not the same problem: a dispute over training material differs from a claim that a particular output copies a work or impersonates a person.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
A broad coalition, not just celebrities
The signatory list spans film, television, music, publishing, comedy, voice work and other creative fields. Alongside widely known names such as Johansson, Blanchett, Kristen Bell, Joseph Gordon-Levitt, Common and Cyndi Lauper, it includes musicians and writers such as Bonnie Raitt, Chaka Khan, Billy Corgan, Jason Isbell, LeAnn Rimes, Jodi Picoult, Jonathan Franzen and George Saunders. The campaign lists R.E.M. members or associated performers individually.
#1 Best Overall
Contemporary coverage reported more than 700 signatories at the campaign’s January launch; that is a dated count, not necessarily the current total. The campaign’s site carries its statement and signatory list. Not every signer necessarily owns all rights to their work: contracts and arrangements with publishers, labels, studios, employers, estates or other rights holders can determine who may authorize particular uses.
Why “theft” is a contested description
The campaign says creative work has been used without authorization and “without regard for copyright law,” and calls the alleged large-scale use “theft.” That language expresses the signatories’ position. It should not be read as a court ruling that every use of copyrighted material to train an AI system is illegal. The legality of a particular use depends on the facts, the material and the jurisdiction; broad questions about AI training and copyright remain contested.
Several distinct issues are often compressed into the phrase “AI training”:
Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Clear out junk files and repair common Windows errorsFree Scan →- Training-data access: what works were collected or copied, how they were obtained, and whether permission or another legal basis applied.
- Output copying or memorization: whether a system reproduces protected expression from a particular work. This is separate from the question of what material was used in training.
- Style imitation: a request for work “in the style of” an artist raises different questions from copying a specific protected work. Style, ideas and artistic influence do not map neatly onto copyright protection.
- Voice and likeness: using a recognizable voice or face can raise publicity, contract, false-endorsement or other legal issues in addition to copyright.
- Economic displacement and fraud: concerns about lost work, synthetic impersonation or deepfakes may be serious without being identical to a copyright claim.
For that reason, the campaign’s accusation should be attributed rather than turned into a blanket statement that AI companies have legally stolen every work used in training.
Rank #3
Johansson’s voice dispute shows a separate concern
Scarlett Johansson’s participation is especially notable because she previously said OpenAI’s “Sky” voice for ChatGPT sounded too similar to her own after she had declined to license her voice. OpenAI denied that the voice was intended to imitate Johansson and said it had been cast before the company contacted her representatives, according to contemporary reporting. The dispute illustrates why performers may see AI as an issue of identity and consent as well as a question about training material; it does not resolve the legal claims surrounding the campaign.
Licensing could help, but it is not simple
Licensing offers a route to documented permission, compensation and potentially clearer provenance for training material. It could also reduce uncertainty by giving AI companies and rights holders a negotiated framework. But building one is difficult: vast collections may contain millions of works, rights can be split among multiple parties, and smaller creators may have less bargaining power than large publishers or studios. A blanket deal may not reflect the value of individual works, while clearing public-domain, user-generated, foreign or orphaned material can raise further complications.
Licensing is not the only model under discussion. Opt-in datasets, opt-out systems, collective licensing, revenue sharing, rights metadata, dataset documentation and safeguards against digital impersonation are among possible approaches. None is established here as a universal fix. The campaign’s demand is a direction—authorization and negotiated relationships—rather than a complete technical or legal blueprint.
This is advocacy, not a lawsuit
The campaign is an open public appeal, not, on the available evidence, a lawsuit or legislative filing. Signing it does not itself initiate litigation, and the appeal does not amount to an order requiring AI companies to stop developing systems. Its significance is that a large and varied group of creators is pressing for consent, compensation and a greater say in how their work is used.
Read the campaign’s statement and signatory list alongside coverage of its January 22, 2026 launch. The central argument is not that AI must disappear, but that its development should not depend on treating human creative work as an uncompensated resource.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
