Robby Starbuck’s lawsuit against Google raises questions about AI accountability. As companies brand AI as “intelligent,” they face higher expectations for accuracy. If AI makes false, damaging claims, responsibility may shift from platforms to creators or publishers. The case could set new legal standards for AI liability and highlights the risks of promising “intelligence” in technology.
The Case That Could Redefine AI Accountability
When conservative activist Robby Starbuck filed a defamation suit against Google this month, alleging its AI systems falsely linked him to sexual assault and white nationalism, it wasn’t just another skirmish in the misinformation wars. It was a test of what happens when a brand markets its technology as intelligent and that “intelligence” misfires.
According to The Wall Street Journal, Starbuck’s complaint alleges that Google’s Gemini and Gemma AI models generated factually incorrect and reputationally damaging statements. This is his second AI-related defamation suit in a year; he previously sued Meta over similar hallucinations, later reaching a private settlement.
No U.S. court has yet awarded damages for AI-generated defamation, but this case could move the law closer to defining where responsibility ends for a tool and begins for its maker.
Section 230 Meets Its Limit
For decades, tech platforms have relied on Section 230, the statute that shields them from liability for user-generated content. But AI changes that equation. When a chatbot creates the content not merely hosts it—the question becomes: Who is the speaker?
Disclaimers like “AI may produce inaccurate information” aren’t likely to carry much weight once a company is on notice that its system fabricates falsehoods about a person and fails to correct them. The more “intelligent” the brand claims its system is, the higher the expectation that it will get the facts right.
Branding “Intelligence” Comes With Responsibility
In the rush to market artificial intelligence, companies like Google, Microsoft, and OpenAI have spent billions promoting the idea that their technology doesn’t just process information, it understands it. They could have positioned these systems as “smart tools,” “efficient assistants,” or “advanced search engines.” Instead, they chose the more evocative and riskier word: intelligent.
That’s not semantics. It’s a promise.
A brand is, at its core, a promise to its users a pledge that it will deliver consistently on what it claims to be. “Smart” suggests helpfulness, efficiency, and responsiveness. “Intelligent,” on the other hand, promises discernment, judgment, even wisdom.
By branding a product as intelligent, a company invites users to trust its reasoning, not just its programming. And in the courtroom, that brand promise becomes evidence of expectation. You can’t market human-level understanding one day and plead mechanical neutrality the next.
From Product to Editor
Whether Google likes it or not, its AI systems now behave like editors deciding which facts to surface, which sources to trust, and which connections to draw. Those aren’t mechanical functions; they are editorial acts.
In branding terms, Google has crossed from platform to publisher. That positioning helped make AI seem accessible and human, but it also carries the burden of human-level accountability.
If an AI tool presents fabricated claims as fact, courts and consumers alike will ask: Who was responsible for accuracy? In journalism, that answer is the editor. In AI, it may soon be the brand behind the machine.
The Emerging Standard of Care
Starbucks’ case may not set a binding precedent, but it will accelerate the debate over what constitutes reasonable diligence for AI companies that brand their systems as “intelligent.” At a minimum, it highlights the need for internal correction protocols once false outputs are identified. More broadly, it challenges companies to define editorial oversight within systems that increasingly act as autonomous communicators.
Just as news organizations need to have developed standards for sourcing and retraction, AI developers will have to establish their own rules of factual accountability. Anything less risks transforming “artificial intelligence” into “artificial negligence.”
The Power and Price of the Word “Intelligent”
Marketers know that words matter. “Intelligent” carries emotional and cognitive weight it signals not just capability, but credibility. That’s why it sells. But it also raises the bar.
Once you call something intelligent, you’ve made a brand promise that goes far beyond performance. You’ve implied trustworthiness. And in an era when machines are writing headlines, answering legal questions, and generating news summaries, that promise may soon be tested under oath.
The smarter move, it turns out, may have been to call these systems “smart.”
Because once you call them “intelligent,” you’re no longer just in the technology business.
You’re in the truth business and in the truth business, there’s no such thing as a harmless hallucination.
This article originally appeared in Law.com and The New York Law Journal.