Photo by Markus Winkler on Unsplash
What happens when we can bring the dead back to life with AI? For Zelda Williams, daughter of the late Robin Williams, the answer isn’t awe or inspiration. It’s grief, frustration, and a heartfelt plea: stop.
In a direct and emotional Instagram story, Zelda didn’t mince words. “Please, just stop sending me AI videos of Dad,” she wrote. “Stop believing I wanna see it or that I’ll understand. I don’t and I won’t.”
And honestly, who can blame her?
The Rise of Sora and the Deepfake Dilemma
The timing of Zelda’s post isn’t random. It comes just days after OpenAI launched Sora 2 — an upgraded video model — alongside a new social app simply called Sora.
Sora lets users create unnervingly realistic AI-generated videos, including digital versions of themselves, friends (with permission), or… dead people.
Photo by Kris Møklebust on Unsplash
While Sora has limits in place to prevent users from generating deepfakes of living people without consent, the rules pretty much vanish when it comes to the deceased.
Dead celebrities from Robin Williams to Martin Luther King Jr. are suddenly popping up in user-generated clips, dancing across screens and saying things they never actually said.
Wait — Isn’t That Illegal?
Surprisingly, no. You can’t legally libel a dead person.
According to existing U.S. defamation laws, libel only applies to the living. And that’s part of the problem. There’s a legal gray zone when it comes to using someone’s posthumous likeness, especially if their estate hasn’t licensed that use.
Photo by Google DeepMind on Unsplash
For companies like OpenAI, this creates a loophole big enough to drive an AI-powered truck through.
OpenAI hasn’t replied to media inquiries about its stance on deepfaking the deceased. But TechCrunch testing suggests that Sora can generate videos of Robin Williams, even though it won’t create ones of Jimmy Carter or Michael Jackson. The logic behind those distinctions is unclear.
“To Watch the Legacies Be Condensed…”
Zelda’s frustration is about more than just her father. It’s about humanity.
“To watch the legacies of real people be condensed down to ‘this vaguely looks and sounds like them so that’s enough,’ just so other people can churn out horrible TikTok slop puppeteering them is maddening,” she wrote.
Her words hit hard because they’re not just about one man. They’re about what it means to respect someone’s memory — and what we lose when we don’t.
Why This Matters
Sora isn’t the only tool out there. Other platforms with fewer restrictions can already be used to generate AI deepfakes that cross ethical and legal lines — from impersonating real people to creating explicit content.
But Sora might be the most accessible, realistic, and potentially popular deepfake tool we’ve seen yet. Which is why the backlash matters.
CEO Sam Altman originally said Hollywood studios and creators would have to “opt out” if they didn’t want their content included in Sora. That means their likenesses would be fair game unless they said otherwise. After industry pushback, he reversed course, saying Sora will offer more detailed opt-in controls in response to copyright concerns.
But none of this helps the dead. They can’t opt in. They can’t say no. And their families, like Zelda, are left watching strangers reshape their legacies with zero say.
Where We Go From Here
The law may not protect the dead, but that doesn’t mean ethics go out the window. If you’re tempted to create or share a deepfake of a historical figure or beloved celebrity, maybe ask yourself: is this honoring them — or exploiting them?
As AI deepfake technology becomes easier to use and harder to detect, we’re all going to have to make choices about where the line is.
Because just because you can… doesn’t mean you should.
📌 Keywords: AI deepfakes, Robin Williams, Sora AI, OpenAI, AI ethics, Zelda Williams, digital legacy, deepfake controversy
✍️ Written for Yugto.io — real tech talk without the fluff.
