Jack Thorne warns writers using AI to 'cheat' and calls for legal bans

Jack Thorne, who wrote Harry Potter and the Cursed Child and co-wrote the Netflix drama Adolescence, stood in London and made a distinction that he believes could reshape his profession. There is a difference, he said, between using artificial intelligence as a tool—to catch a typo, to organize a scene—and using it as a replacement for the work itself. Some writers, he warned, are doing the latter. They are feeding prompts into an AI system, collecting the output, and passing it off as their own writing. That, Thorne said, is cheating.

Thorne is president of the Writers' Guild of Great Britain, and he spoke with the weight of that position. He had talked to executives and studio heads about their use of AI and heard troubling stories from writers he knows. The problem, as he framed it, is not the technology itself but the dishonesty of its deployment. A writer who generates a script through AI without telling their employer, without declaring it, is committing fraud. They are stealing from the people who hired them and from the writers competing fairly for the same work.

What makes this urgent, in Thorne's view, is that the industry and government have not yet drawn clear lines. He called for laws that would make undisclosed AI use unlawful—that would treat it the way universities treat plagiarism in student essays, with real consequences for a writer's career. Broadcasters like the BBC and Channel 4 should threaten expulsion from their rosters. The impact, he said, needs to be substantial enough to actually deter the behavior.

But Thorne's concern runs deeper than individual misconduct. He pointed to a larger vulnerability. If studios and publishers can point to writers as the problem—if they can say, "See, writers are using AI to cut corners"—then those companies have cover to do whatever they want with AI themselves. They can justify scraping writers' work to train their models, they can justify replacing human writers with systems trained on stolen material. The individual cheater becomes the excuse for systemic theft.

This connects to Thorne's second major complaint: government inaction. The UK government, he said, has been "a bit dazzled by tech." Last year, ministers abandoned a proposal that would have required artists and copyright holders to opt out of having their material fed into AI training systems. Instead, they launched a consultation. Thorne sees this as insufficient. He argued that the government should require AI companies to get permission before using creative work—an opt-in system, not an opt-out. The pieces of AI being built today, he said, cannot be built on what amounts to copyright theft.

Thorne framed the stakes in terms of time. The decisions the government makes in the next five years, he said, will define what the next hundred years look like for writers in Britain. This is not hyperbole in his telling—it is a statement about path dependency, about how early choices in a technology's regulation can lock in patterns that become nearly impossible to change. He called for government leadership, even global leadership, to set clear rules before the industry calcifies around current practices.

A government spokesperson responded by saying that copyright law must protect Britain's position as a creative powerhouse while also unlocking AI's potential. The statement offered no specifics about how those two goals would be balanced, or whether the government would move toward opt-in rather than opt-out protections. It was the kind of response that suggested the government was still in listening mode, not decision mode—exactly what Thorne was criticizing.