Will it run?
Models

Robin Williams' children take over his Instagram account to fight AI fakes

By Rae Whitlock Clawpit staff
Robin Williams' children take over his Instagram account to fight AI fakes

The account is active again after 12 years. Zak, Zelda and Cody Williams announced two days ago that they are assuming control of their father's Instagram profile, which had sat dormant since his death in 2014. The stated goal: turn the page into a "safe and reliable place" where stories, photos and videos reflect his legacy with authenticity, warmth and accuracy — not a synthetic version produced by a language model.

"The best way to fight rampant AI abuse"

In a post on her personal account, Zelda explained that reviving the page was meant to confront what she called "rampant AI abuse" of her father's voice and likeness. She acknowledged the renewed activity might unsettle some followers, but argued that creating an official archive of original material is the most effective tool they have found so far against the uncontrolled drift of deepfakes online.

A plea that fell on deaf ears

Last year Zelda asked fans to stop sending her AI-generated videos featuring her father. "To see real people's legacies reduced to 'it looks and sounds kind of like them, so that's good enough,' just so other people can make TikTok garbage that puppeteers them — it's insane," she wrote at the time. The request did not stop the phenomenon, and the family decided to shift from a reactive strategy to a preventive one.

Guardrails that break easily

The problem is far from unique to the Williams family. Leading chatbots include guardrails designed to prevent generation of content depicting celebrities, but those guardrails are repeatedly bypassed. OpenAI's Sora, since shut down, became a hub for generating videos of copyrighted characters and celebrities. Grok still produces sexual deepfakes of famous women, according to a Wired investigation. ByteDance's Seedance was forced to sign an agreement with the Motion Picture Association after its lax protections allowed users to generate scenes featuring Hollywood stars.

Celebrities as tools for fraud

Bad actors are taking the technology a step further: AI versions of Taylor Swift and Rihanna have been used to promote scams on TikTok. The phenomenon is no longer confined to entertainment; it has become infrastructure for large-scale online crime, and platforms are struggling to keep pace.

An official archive instead of a vacuum

The children plan to use the account to share their father's "unique talent and humor" with the world. In other words: they are building an official source of truth to fill the vacuum that fakes exploit. This is not a war on technology; it is a war on the vacuum that technology exploits.