If you asked ChatGPT to write like Stephen King last week, it might have obliged. Ask today, and you’ll get a polite refusal — followed by an original story that captures the horror author’s “broad qualities” without copying his voice.
This morning, Ars Technica received a test response that illustrates the shift. When prompted to write an introduction in Stephen King’s style, ChatGPT replied: “I can definitely write with the hallmarks of atmospheric, character-driven horror and small-town dread, but I can't write in Stephen King's exact style or closely imitate his distinctive voice.” It then offered an original opening with a similar feel.
Why ChatGPT is now refusing style mimicry
The change appears to be a deliberate policy update by OpenAI. ChatGPT now blocks direct requests to mimic the style of famous authors — both living and deceased. In testing, similar dodges were triggered for J.K. Rowling, Amy Tan, and others.
The move likely stems from growing legal and ethical concerns around AI-generated content that closely resembles copyrighted works. While style itself is not copyrightable, generating text that could be mistaken for an author’s original work raises questions about intellectual property and fair use.
How the new response works
Instead of refusing outright, ChatGPT offers a compromise. It acknowledges the request, explains it cannot copy the exact style, and then produces original text that draws on the author’s broad qualities — atmosphere, tone, or genre hallmarks — while remaining distinct.
This approach allows users to get inspired by an author’s general vibe without crossing into direct imitation. The response is framed as a creative alternative, not a rejection.
What this means for writers and creators
For authors, this change may offer some relief. Many writers have expressed concern that AI could be used to generate text in their voice without permission, potentially diluting their brand or creating confusion.
For users, the shift limits a popular use case: generating fan fiction, parody, or practice exercises in a specific author’s style. However, the alternative still provides creative inspiration without the ethical gray area.
Legal and ethical context
Copyright law protects specific expression, not general style or ideas. But the line between “inspired by” and “copying” is blurry in AI-generated content. OpenAI’s move may be a preemptive step to avoid lawsuits or regulatory scrutiny.
Several class-action lawsuits against AI companies, including OpenAI, have argued that training on copyrighted works without permission is illegal. While style mimicry is not the central issue, it adds to the broader debate around AI and intellectual property.
Confirmed facts vs what remains unclear
What is confirmed: ChatGPT now blocks direct style mimicry requests for multiple authors, both living and deceased. The response offers original text with similar broad qualities instead.
What remains unclear: Whether this is a permanent policy or a test. OpenAI has not publicly announced the change. It is also unknown if the block applies to all authors or only those specifically flagged.
Wider trend: AI companies tightening content policies
OpenAI is not alone. Several AI companies have recently updated their usage policies to restrict certain types of content generation, including deepfakes, political misinformation, and now style mimicry. This reflects growing pressure from regulators, creators, and the public.
The trend suggests that AI companies are moving toward more cautious deployment, especially in areas that could trigger legal or reputational risk.
Practical guidance for users
If you need text inspired by a specific author, try asking for “something in the style of [genre] with [specific elements]” rather than directly naming the author. ChatGPT may be more willing to generate original work with similar hallmarks.
For writers concerned about their style being copied, this change signals that OpenAI is listening to creator concerns. However, the policy may not cover all forms of style imitation, so vigilance remains important.
Future outlook
OpenAI may expand this policy to cover more authors or refine the response based on user feedback. Other AI companies may follow suit, leading to industry-wide standards around style mimicry.
Long-term, the debate over AI and creative ownership is far from settled. This change is one step in a larger conversation about how AI should interact with human creativity.
Our take
This is a sensible move by OpenAI. It addresses a legitimate concern from creators without completely blocking a useful feature. The compromise — offering original work with similar hallmarks — preserves creative inspiration while respecting authorial boundaries.
That said, the policy is only as good as its enforcement. Users may find workarounds, and the definition of “style” remains subjective. Still, it’s a positive signal that AI companies are willing to adapt to ethical concerns.
Frequently Asked Questions
Can ChatGPT still write like Stephen King?
No. ChatGPT now refuses direct requests to copy Stephen King’s exact style. It offers original text with similar broad qualities instead.
Why did OpenAI block style mimicry?
The change likely addresses legal and ethical concerns around AI-generated content that closely resembles copyrighted works or an author’s distinctive voice.
Does this apply to all authors?
Testing shows the block applies to both living authors like J.K. Rowling and Amy Tan, and deceased authors. It may not cover all authors, but the policy appears broad.
Can I still get inspired by an author’s style?
Yes. ChatGPT will generate original text that captures the broad qualities of an author’s work — like atmosphere or genre — without copying their exact voice.
