William Shatner has lived long enough to watch himself become a digital ghost he never authorized. The 94-year-old actor — still sharp, still combative, still unmistakably Shatner — is waging a very public war against artificial intelligence systems that fabricate stories about him and present them as fact. Not exaggerations. Not misquotes. Whole-cloth inventions, delivered with the confident authority of a machine that doesn’t know or care that it’s lying.
The latest salvo came on X, where Shatner posted a pointed complaint about AI chatbots generating false biographical information about him. According to Futurism, the actor has been increasingly vocal about what he describes as AI “spreading rumors” — fabricated details about his personal life, career, and relationships that users encounter when querying popular AI tools. These aren’t deepfake videos or cloned voices, though those problems exist too. This is something more insidious: text-based hallucinations that carry the veneer of encyclopedic reliability.
“AI is spreading rumors about me,” Shatner wrote, distilling a complex technological failure into six blunt words.
He’s right. And the problem extends far beyond one nonagenarian celebrity.
Large language models — the technology underpinning ChatGPT, Google’s Gemini, Meta’s Llama, and others — don’t retrieve facts the way a search engine does. They predict the next most likely word in a sequence based on statistical patterns in their training data. When the data is thin, contradictory, or absent, the models improvise. The AI industry has a clinical term for this: hallucination. Critics argue the word is too generous. The machines aren’t hallucinating. They’re confabulating — filling gaps with plausible-sounding fiction and presenting it without caveat or uncertainty.
For public figures like Shatner, this creates a specific and personal kind of damage. Ask an AI chatbot about a celebrity’s marriages, feuds, political views, or health, and you may receive an answer that blends real facts with invented ones. The user has no reliable way to distinguish between the two. The output looks authoritative. It reads like a Wikipedia entry. But it might be partly or entirely wrong.
Shatner’s frustration isn’t theoretical. He has pointed to specific instances where AI systems generated false claims about his life. The details of those claims vary — some involve fabricated quotes, others involve invented personal history — but the pattern is consistent. The machines produce misinformation about a real, living person, and that misinformation circulates as though it were verified.
This isn’t a new concern for Shatner. He has been one of the most digitally engaged nonagenarians on the planet, maintaining an active and often combative presence on X for years. He’s tangled with trolls, corrected journalists, and pushed back against misinformation long before generative AI became a mainstream consumer product. But the scale of the problem has changed. A human troll can be blocked. A rumor on a message board can be debunked. An AI system that generates false information on demand, millions of times a day, across dozens of platforms? That’s a different order of magnitude.
The entertainment industry has been grappling with AI-related threats for more than two years now, ever since generative AI tools began producing convincing synthetic media at scale. The 2023 SAG-AFTRA strike was driven in part by fears that studios would use AI to replicate actors’ likenesses without consent or compensation. The resulting contract included provisions around AI use, but enforcement remains uneven, and the technology has continued to advance faster than the legal frameworks meant to contain it.
Shatner’s complaint touches a different but related nerve. It’s not about his likeness being used in a film without permission. It’s about his identity being corrupted by machines that generate false narratives about who he is. The distinction matters. Deepfake protections focus on visual and audio replication. But what about textual fabrication? What legal recourse does a person have when an AI system tells millions of users something untrue about them — not as satire, not as opinion, but as apparent fact?
The answer, right now, is: very little.
U.S. defamation law requires proving that a false statement was made with actual malice or negligence, depending on the plaintiff’s public-figure status. Applying that standard to an AI system is legally murky at best. The AI didn’t “intend” anything. It doesn’t have malice. It’s a statistical model producing outputs based on probability distributions. Courts haven’t yet established clear precedent for holding AI companies liable for hallucinated defamation, though several cases are working their way through the system.
Meanwhile, the companies building these models have largely treated hallucinations as a technical problem to be solved through better training and reinforcement learning, not as a legal or ethical crisis requiring immediate structural change. OpenAI, Google, and others have added disclaimers noting that their systems can make mistakes. But disclaimers don’t undo the damage when a false claim about a real person enters the information supply chain and gets repeated, shared, and cited as though it were true.
Shatner’s case is illustrative precisely because of who he is. He’s famous enough that most people have some baseline knowledge of his real biography — Captain Kirk, the Priceline commercials, the spoken-word albums, the trip to space on Blue Origin in 2021 at age 90. That fame provides a partial check against AI fabrication; people are more likely to notice when a chatbot says something obviously wrong about William Shatner than about a less well-known individual. But for the millions of private citizens, minor public figures, academics, small business owners, and local politicians who lack that level of public recognition, AI hallucinations can become the de facto record. No one is checking. No one is correcting.
And the problem is compounding. As AI-generated text proliferates across the web, it becomes training data for the next generation of models. A hallucination generated today can become an entrenched “fact” in tomorrow’s training corpus. Researchers call this model collapse — the degradation that occurs when AI systems train on AI-generated content. It’s a feedback loop with no natural correction mechanism.
Shatner, to his credit, isn’t just complaining. He’s using his platform to draw attention to a systemic failure that affects everyone. His posts on X have generated significant engagement, with thousands of users sharing their own experiences of AI systems producing false information about them or people they know. The anecdotal evidence is overwhelming. The formal accountability structures are not.
Some legislative efforts are underway. The EU’s AI Act, which began phased implementation in 2024, includes transparency requirements for general-purpose AI systems, including obligations to disclose when content is AI-generated and to take steps to reduce the generation of false information. In the United States, progress has been slower. Several states have passed or proposed laws targeting AI-generated deepfakes, particularly in the context of elections and pornography, but comprehensive federal legislation addressing AI hallucinations and textual fabrication remains elusive.
The tech industry’s position has been that these problems will diminish as models improve. There’s some truth to that — newer models hallucinate less frequently than their predecessors, and techniques like retrieval-augmented generation, which grounds AI responses in verified source material, have shown promise. But “less frequently” is not “never,” and the sheer volume of AI interactions means that even a small hallucination rate produces an enormous absolute number of false claims.
Shatner isn’t waiting for the technology to catch up. He’s 94. He’s watched the internet evolve from novelty to necessity, watched social media go from promise to pathology. He’s seen this pattern before — a powerful new technology deployed at scale before anyone has figured out the rules. The difference this time is that the technology doesn’t just amplify human speech. It generates its own. And it does so with a fluency and confidence that makes its errors nearly indistinguishable from its truths.
“The machines are making things up about me,” is a sentence that would have sounded absurd a decade ago. Now it’s just Tuesday.
For the AI industry, Shatner’s complaints should register as more than a celebrity grievance. They’re a signal. When a 94-year-old man with no particular technical expertise can identify and articulate the core failure mode of your product — that it lies, fluently and without remorse — you have a trust problem that no amount of fine-tuning will fully resolve. The fix isn’t just technical. It’s structural, legal, and cultural. And it needs to happen before the next generation of models makes the current ones look quaint.
Shatner has spent six decades playing characters who confront the unknown. This time, the unknown is speaking in his voice, telling stories he never lived, and millions of people are listening.
WebProNews is an iEntry Publication