Scottish National Party MP Pete Wishart’s fury at Elon Musk’s chatbot Grok has become more than a local defamation row. It is a test case in the fast-approaching clash between artificial intelligence and free expression.

When Grok, hosted on Musk’s social platform X, falsely described Wishart as a: “rape enabler,” he demanded that the AI be: “recalibrated or shut down.” The post, quickly retracted, triggered threats against him and a storm of commentary about the power of automated systems to amplify libel.

The United Kingdom, already drafting its Online Safety Act regulations, now faces pressure to expand those rules to cover generative AI. “This is deeply distressing,” Wishart said, calling the exchange “appalling and defamatory.” Data-ethics experts responded that such outputs are not “opinions” but algorithmic reproductions of public discourse, yet once they appear on mass platforms they acquire the force of speech.

Free-speech advocates may counter that forcing Grok to limit its responses would hand governments the precedent to decide what an AI may or may not say. Musk’s xAI team, which markets Grok as answering: “spicy questions rejected by most systems,” argues that over-regulation will sterilise public debate and concentrate truth in the hands of regulators. In liberal democracies built on open discourse, the principle at stake is whether flawed but transparent machines should have their speech rights limited by governments.