ChatGPT is changing how it responds when users ask it to write like a famous author. Instead of directly copying a recognizable literary voice, the system now says it can draw on broader creative qualities while keeping the output distinct.
That may sound like a small wording change. But for writers, publishers, AI companies and users, the difference between imitating a specific voice and producing something only generally similar is becoming increasingly important.
What Changed In ChatGPT
According to Ars Technica, OpenAI's ChatGPT is now refusing prompts that ask for text in the exact style of well-known authors. In a test involving Stephen King, ChatGPT did not agree to closely imitate his voice. Instead, it offered to write with elements such as atmospheric, character-driven horror and small-town dread while keeping the work original.
Ars Technica also tested prompts involving other authors. The refusals appeared for living writers including J.K. Rowling and Amy Tan, as well as deceased writers including Charles Dickens and Ernest Hemingway.
The pattern is not simply a refusal to help with writing. ChatGPT still offers a substitute: it can produce an original passage that uses broad traits associated with a genre, tone or literary tradition. The key point is that the model avoids saying it will reproduce an author's exact style or distinctive voice.
Why The Wording Matters
The distinction between exact imitation and general inspiration may become important because OpenAI is already facing a number of lawsuits from book authors. Those authors allege large-scale copyright infringement connected to models trained on their work.
One lawsuit cited ChatGPT's "uncanny ability to generate text similar to that found in copyrighted textual materials." That concern sits close to the issue raised by style imitation: if a model can produce work that feels highly similar to a specific writer, critics may argue that the output affects the value or control of the original work.
In the US, copyright law generally protects a particular expression of an idea rather than an author's more abstract style. But the Ars Technica article notes an important legal boundary: an AI-generated imitation could become infringing if it is "substantially similar" to the original author's work.
That makes ChatGPT's new response strategy notable. By redirecting users toward broad qualities rather than exact mimicry, the model appears to reduce the chance that a prompt explicitly asks for a close copy of a named writer's distinctive voice.
Writers Are Being Warned Too
The Authors Guild has also urged caution around generative AI and style mimicry. In a “best practices” document, the organization advises writers not to use generative AI to purposely copy or mimic the unique styles, voices or other distinctive attributes of other writers' works in ways that harm the value of those works or attempt to profit from them.
The Authors Guild also warns that, beyond ethical concerns, mimicking another writer's voice or style could lead to claims of unfair competition or copyright infringement.
That advice points to a practical concern for professional writing. A user may see a style prompt as a shortcut for tone, pacing or genre. But if the prompt is built around a named author, the result can raise questions about whether the output is original writing or an attempt to benefit from someone else's recognizable creative identity.
Other AI Tools Handle Style Requests Differently
Not every major AI model treats these prompts the same way. A No Latency study found that Google's Gemini consistently complied with requests to copy an author's style. Perplexity AI's models consistently refused such requests and redirected them in a way similar to ChatGPT.
The same study placed Anthropic's Claude and Microsoft's Copilot somewhere between those approaches. Those models complied with style-mimicry requests but added qualifications that suggested, according to the authors, "an awareness of the imitation issue."
OpenAI has already made a clearer public rule in image generation. The company says its DALL·E 3 model is designed to decline requests asking for an image in the style of a living artist. Ars Technica notes, however, that an extensive OpenAI model spec published in December did not include a clearly stated ban on mimicking style in written responses.
That contrast makes the ChatGPT behavior more significant. Even if the written policy language was not spelled out in the same way, the product behavior now appears to be moving toward stronger limits on direct author-style requests.
What This Means For Users
For people who use ChatGPT as a writing assistant, the change may require different prompts. A direct request to write like a named author may now be blocked or redirected. A request for broader creative attributes may still work.
That means users may need to describe the effect they want instead of naming the person they want to imitate. For example, they can ask for a passage with suspense, concise prose, small-town tension, comic timing or formal narration without asking the model to copy a specific author's voice.
Some longtime users are already frustrated. Ars Technica cited a Reddit user, Dazzling-Major-5620, who wrote that ChatGPT now says it cannot generate content in the style of specific authors and wondered how to work around the change.
The broader direction is clear from the reported behavior: ChatGPT is still willing to help with style, tone and genre, but it is drawing a sharper line around direct imitation of famous writers. As legal pressure and professional norms evolve, that line may become a central feature of how AI writing tools are used.