Nicolas Cage told a New York Comic Con audience on Saturday, October 10, 2026 that he did not sign an AI waiver for Prime Video's Spider-Noir, and that he will "probably" not work with Amazon's studio again. Variety reported the remarks, and several outlets including TheWrap and Deadline carried them too. This is a single on-stage account from one party. Amazon has not published the waiver or responded in the coverage we found. Below is what he said, what is verified, and what readers who sign AI-related contracts can take from it.
Claimed vs verified at a glance
| Claim | Status |
|---|---|
| Cage said he refused to sign an AI waiver for Spider-Noir | Verified as a statement: quoted by Variety and others |
| Amazon asked him to sign such a waiver | Cage's account only; Variety says "allegedly" |
| The waiver's exact terms | Not published |
| Amazon has invested $50 billion in OpenAI | Cage's figure; we have not verified it, and our own coverage of OpenAI's finances shows numbers in this area get muddled |
| Spider-Noir was cancelled after one season | Reported by Variety, about a month before the panel |
| The refusal caused the cancellation | No evidence; reported budget reasons |
What exactly did Cage say?
According to Variety, Cage said: "Amazon is a very AI-friendly company, to the tune of $50 billion invested in OpenAI. I am not an AI-friendly actor, so it's like, connect those dots! I'm not a member of their club. I didn't sign the waiver to let him use AI on 'Spider Noir.' I'm probably not going to be working with that studio again, but that's okay."
Variety adds context: Amazon cancelled the superhero series after one season last month, days before the Emmy weekend, where it had 11 nominations. The show reportedly drew 2.6 billion viewing minutes in its first six weeks. The same panel produced a tease that Cage's next project is an HBO True Detective season, which HBO had not confirmed.
What we do not know
We do not know what the waiver said. "AI waiver" could mean many things: consent to scan and store a performer's likeness, to use AI tools in post-production, to train models on footage, or to create digital replicas. Each has very different consequences. Cage's remarks do not specify, and reporting has not shown the paperwork.
We also do not know whether the request was unusual. Some studios now put AI language in standard contracts, so a clause existing does not mean it was aimed at one actor. Without the document, any claim that Amazon targeted him, or that the refusal affected the cancellation, is speculation. Cage himself did not say the waiver was the reason for the cancellation as quoted.
Why an actor's refusal matters: the background
Cage has been vocal on this before. In 2024, Variety's Australian edition reported on his fear of AI and the digital body scans he gave for projects, including Spider-Noir, and his hope they would not be used to recreate him after his death. That earlier interview is context, and it shows a long-running position rather than a one-off remark.
Performer consent has been a live legal area. New York's law on AI video disclosure and synthetic performers is one example of lawmakers moving here. Our coverage of the first MPA-rated AI movie shows the other side, where studios release fully AI-made films. Between those poles sit the contracts most actors actually sign.
How this fits the wider fight over performer consent
The 2023 Hollywood actors' strike put digital replicas at the center of bargaining. The SAG-AFTRA agreement that ended it added rules requiring consent and compensation when a studio creates or uses a performer's digital replica, and separate rules for synthetic performers. Those protections apply to union productions and set a floor, but a floor is not a ceiling. Individual contracts still carry clauses, and stars with leverage can push back harder than supporting actors who may feel they cannot refuse.
That is part of why Cage's account resonates. A well-known actor saying no is a visible act, and he can afford the consequences. A working actor on a first credit may not be able to. Whether Cage's refusal was unusual or routine on this production is exactly the detail missing from public reporting.
There is also a technical reason these clauses keep getting broader. Generative video and voice tools now make it cheap to alter a shot after the fact, so studios want blanket permission rather than negotiating each use. Performers, in turn, worry that a scan taken for one purpose, such as a visual effects pass, becomes raw material for something else. Cage voiced that exact worry in 2024 about his scans surviving him.
What this means for creative professionals and for AI buyers
For creative professionals the lesson is practical: the AI language in a contract is now a negotiating item, not boilerplate to skim. Ask what happens to captured data, who can access it, whether it can be used to train a model, and what happens at the end of the contract.
For companies that buy or deploy AI, the story is a reminder that consent and provenance are becoming reputational and legal risks. A studio, publisher or agency that cannot say exactly what it may do with a person's likeness will face questions, from talent and from regulators. The disclosure rules in New York and the music chart restrictions in Australia show governments already treating synthetic media as something to label and limit.
For readers following AI news more generally, treat on-stage claims as a starting point. The useful follow-ups are the waiver's text, any statement from Amazon, and whether other performers on the series report the same request. Until those appear, the verified facts are narrow: Cage said it, outlets quoted him, and the cancellation happened. The rest is interpretation.
How AI clauses in performer contracts work
Language varies, but typical elements include:
- Scan and capture consent. Whether your face, body or voice may be digitally captured.
- Digital replica rights. Whether the production may create a synthetic version, for which scenes, and for how long.
- Training rights. Whether recordings may be used to train models, a separate question from using a replica.
- Post-production edits. Using AI to change dialogue, de-age, or fix shots.
- Compensation and approval. Whether each use is paid and whether you approve it.
- Duration and death. Whether permission survives the contract and your lifetime.
The practical advice for any performer, and for freelancers in other fields facing similar clauses: ask for the AI language in plain terms, ask whether it is negotiable, and have a union representative or lawyer review it before signing. Our piece on how publishers are using AI for covers and copy shows creative workers elsewhere raising the same disclosure concerns.
Why the Amazon and OpenAI link is the interesting part
Cage's framing is that a company investing heavily in AI will want contracts that let it use AI. That is an inference, and it is a common one among creative workers. We could not confirm the $50 billion figure from the sources we read, so treat it as his number. If you are reporting on this, ask Amazon or check filings.
The broader point holds regardless: as AI developers and media companies get closer, contract language on likeness becomes a bargaining front. Workers who decline are making a business choice that can carry costs, which Cage acknowledged by saying it was fine if he never worked there again.
What would change our assessment
Three developments would move this from a claim to a confirmed story. First, Amazon or Sony Pictures Television publishing or describing the waiver language, which would show whether it covered scans, post-production edits or model training. Second, other cast or crew confirming they were asked to sign the same document, which would show whether this was standard paperwork. Third, a union statement, since guild rules govern what a production may require of a member. We will add an update section here if any of these arrive.
How to check claims like this
- Look for the original recording or a verbatim transcript, not just paraphrases. Variety quotes him in full.
- Look for a response from the company. None was reported at the time of writing.
- Separate the cancellation timeline from the waiver claim. They are different facts.
- Ask for documents. The waiver text is the thing that would settle the scope.
What to do if you are asked to sign an AI waiver
If you handle your own contracts, whether as an actor, voice artist, writer or consultant, read any clause that mentions AI, machine learning, synthetic media, digital replicas or training. If you build or deploy AI agents that act on contracts or personal data, the same principle applies: constrain what the system can do before it acts. AgentBeam is the agent security platform from the explainx.ai team that stops AI agents before they take dangerous actions.
Related reading
- New York AI video disclosure law and synthetic performers
- First MPA-rated AI movie, December 4, 2026
- Big Five publishers quietly using AI for covers and copy
- OpenAI $50 billion run rate: claimed vs verified
- Nikon Small World winner disqualified over AI-generated video
- Australia bans AI music from charts
Based on press reports as of October 10, 2026; Amazon may respond or publish details later.