OpenAI's ChatGPT Now Refuses to Directly Mimic Famous Authors' Writing Styles

Summary

OpenAI has implemented a policy change in ChatGPT that prevents the AI model from directly copying the writing styles of famous authors, particularly those who are still living. When users request text written in the style of authors like Stephen King, J.K. Rowling, or Amy Tan, ChatGPT now declines and instead offers to capture the "broad qualities" of those authors while maintaining its own voice.

The shift appears directly connected to ongoing copyright infringement lawsuits filed by authors against OpenAI, who allege that ChatGPT was trained on copyrighted material and can now reproduce work "substantially similar" to originals. By refusing exact style imitation, OpenAI may be attempting to create legal separation between inspiration and infringement—a distinction that could prove important as these cases proceed through court.

Interestingly, testing by No Latency found that ChatGPT still complies with style-copying requests for deceased authors, suggesting the policy targets living creators whose rights are actively protected. The change mirrors OpenAI's approach with DALL·E 3, which already declines image generation requests in the style of living artists. However, other LLMs have not adopted similar restrictions, and OpenAI's official model specifications lack explicit prohibitions against style replication.

• ChatGPT still complies with style-copying requests for deceased authors, revealing the distinction is based on active intellectual property rights
• Competing LLMs like Google's Gemini have not adopted similar restrictions, creating potential competitive and legal differentiation