Netflix’s new Wonka show has run into backlash after using an AI-generated Gene Wilder voice. Wilder, who died in 2016, appears in the series with the consent of his estate, according to BBC News.

That single detail matters because it cuts against the loudest version of the internet story. This is not a case of a studio allegedly trying to bypass rights holders. The estate gave consent. So the backlash is less about whether consent exists and more about how it is perceived, how it is communicated, and what audiences expect when technology recreates a real person’s likeness and voice.

For executives, this is the emerging fault line in AI media: legal permission and audience acceptance are not the same thing. Copyright and publicity rights frameworks typically focus on ownership and licensing. But the second-order risk is reputational, not just legal. When viewers feel the AI representation is misleading, exploitative, or overconfident, the controversy can spread faster than any clarification. And the faster the spread, the harder it is for brands to contain it to a single headline.

AI voice generation also raises a governance question boards and senior management are increasingly facing. Even when you have permission, you still need a process for deciding how to use the technology and how to disclose it. The BBC’s reporting highlights the estate consent, which implies a formal rights pathway was followed. But audiences are still asking the operational questions: Was the voice clearly labeled as AI? Were creative teams transparent about the production pipeline? Did marketing frame it as an authentic Wilder performance or as an AI recreation? Those details are not included in the BBC excerpt, but they are exactly the type of perception gap that turns a rights win into a trust loss.

This backlash also lands in a broader regulatory environment where governments and regulators are actively trying to catch up to synthetic media. Across jurisdictions, regulators are grappling with how to treat generated likeness and voice, especially when the subject is deceased. Some regimes focus on consent and rights management, while others consider deception and consumer protection. Even if an AI use clears rights hurdles, it can still run into scrutiny if audiences believe they were manipulated or misled. In other words, “estate consent” may help with one category of risk while leaving another wide open.

Now zoom out to incentives. Streaming platforms are built to reduce friction: faster production cycles, lower costs, and more control over creative execution. AI tools can be tempting because they compress time and expand what is feasible. But the Wonka backlash is a reminder that speed and feasibility do not automatically create legitimacy. Viewers do not experience AI as a set of tools. They experience it as a performance from a person they recognize. When the person is dead, the emotional stakes intensify, because the performance cannot be verified or renegotiated the way a living actor’s participation can.

That emotional intensity is why these controversies tend to become board-level issues. Investors and boards care about subscriber trust, churn risk, and long-term brand health. A single project can become a pattern in the public mind: “They use AI voices like this now, so what next?” Once the question becomes “what next,” it stops being a PR problem and becomes a strategic one. Executives at other studios, networks, and platforms with AI ambitions now have a live benchmark for how quickly consent can stop being persuasive.

The practical stakes for peers are clear. If you are considering AI-generated voices, the estate permission highlighted in this case may be necessary but not sufficient. Your internal policy should treat audience communication as part of compliance, not as an afterthought. The Wonka story shows that even with consent in place, backlash can still attach itself to the project, forcing teams to spend time, attention, and credibility on damage control.

In short, the BBC reporting ties the Gene Wilder AI controversy to consent from Wilder’s estate, and that should shape how executives interpret the situation. The question is no longer only “Did we have permission?” It is “Did we earn trust?” And in a world where AI can mimic a recognizable voice, trust can become the scarce resource that decides whether a creative bet pays off or turns into a reputational hangover.