ChatGPT stops mimicking famous authors' styles as OpenAI faces mounting copyright lawsuits

ChatGPT Refuses Requests to Copy Author's Style in Major Policy Shift

OpenAI has quietly implemented a significant change to how ChatGPT handles requests to mimic writing styles of famous authors. The chatbot now declines to generate text that directly copies the distinctive voices of both living and deceased writers, instead offering to capture "broad qualities" while maintaining its own distinct voice 1

1

. When asked to write in Stephen King's style, ChatGPT responds that it "can definitely write with the hallmarks of atmospheric, character-driven horror and small-town dread" but cannot write in King's "exact style or closely imitate his distinctive voice" 1

1

.

This OpenAI policy change marks a notable departure from previous behavior. Testing by multiple outlets found ChatGPT generated similar responses for authors ranging from J.K. Rowling and Amy Tan to Charles Dickens and Ernest Hemingway 1

1

. The shift appears to have occurred rapidly—a study published by No Latency on July 15 found ChatGPT still complied with requests for deceased authors, but by July 27, that gap had closed entirely 3

3

.

Copyright Infringement Concerns Drive Legal Strategy

The timing of this change appears directly connected to OpenAI's mounting legal challenges. The company faces multiple AI copyright lawsuits from authors and publishers alleging large-scale copyright infringement by models trained on their work without permission 1

1

. One lawsuit specifically cites ChatGPT's "uncanny ability to generate text similar to that found in copyrighted textual materials" 1

1

. Prominent authors including George R.R. Martin and John Grisham have joined class-action lawsuits targeting how AI models were trained on pirated e-book libraries 4

4

.

The legal distinction between style and expression matters significantly here. US copyright law generally protects only specific expression of an idea, not the more intangible author writing style. However, AI-generated content and intellectual property law intersect in complex ways—an AI-generated stylistic imitation could become infringing if it grows "substantially similar" to the original author's work 1

1

. Robert Brauneis, a law professor at George Washington University, noted that "we've never had a situation in which this personal style of individual creators could be imitated as well and as inexpensively as we now have with AI" 1

1

.

The shift to mitigate legal risks follows Anthropic's recent $1.5 billion settlement with authors—the largest copyright settlement in US history—which likely prompted OpenAI to reconsider its approach 5

5

.

How Other AI Models Handle Author Imitation Requests

The No Latency audit revealed stark differences in how major AI platforms approach these requests, highlighting that this is a policy decision rather than a technical limitation. Google Gemini consistently complied with requests to imitate the writing style of named authors without any restrictions or disclaimers 3

3

. Anthropic Claude and Microsoft's Copilot took a middle approach, complying with style-mimicry requests but attaching caveats about originality that suggest "an awareness of the imitation issue" 1

1

. Perplexity AI enforces the strictest ban, consistently refusing such requests and redirecting them similar to ChatGPT's new approach 1

1

.

OpenAI has precedent for such restrictions in creative writing tools. Its DALL·E 3 image model already "is designed to decline requests that ask for an image in the style of a living artist" 1

1

. However, an extensive OpenAI model spec published in December didn't include clearly stated prohibitions against replicating copyrighted material or mimicking style in written responses, suggesting this text-based change represents a quiet retrofit rather than stated policy 3

3

.

User Backlash and Practical Implications for Writers

The change has already generated user backlash from writers who relied on this functionality. One Reddit user complained, "Now Ms. GPT says she can't generate content in the style of specific authors. My prompts were soooo specific and I got exactly what I wanted out of them... I have no idea how to get around this other than maybe feed it prose I already edited that's technically mine?" 1

1

. Some users noted they had paid for subscriptions and felt the change disrupted their book-writing process 5

5

.

The Authors Guild, representing professional writers, has taken a firm stance on AI ethics in this area. The organization urges writers to "respect your fellow authors and do not 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 their works or attempt to profit from them" 1

1

. They warn that mimicking a writer's unique voice could subject users to claims of unfair competition or copyright infringement 1

1

.

The issue gained prominence when author Lena McDonald faced widespread criticism after her book included an apparent AI-generated response reading, "I've rewritten the passage to align more with [fellow genre author] J. Bree's style..." 1

1

. Users on Reddit have suggested workarounds, including manually describing author traits rather than naming them, or using competing models like Claude and Gemini that still comply with such requests 5

5

.

What This Means for Training Data and Intellectual Property Rights

This policy adjustment represents OpenAI's attempt to navigate the tension between maintaining a useful creative writing tool while narrowing the specific use that most invites litigation. The company's approach suggests it's trying to draw a line between offering stylistically aligned outputs based on general literary techniques versus producing work that could be deemed substantially similar to copyrighted material. The distinction between capturing a "feeling" versus an "exact style" may seem subtle, but it could prove legally significant as courts grapple with how intellectual property rights apply to AI-generated content 1

1

.

As AI companies face increasing pressure over training data practices and how they handle copyrighted works, expect more policy changes across the industry. OpenAI's licensing deals, such as its pact with Getty Images, point toward one potential resolution, while this type of usage restriction represents another, less expensive approach 3

3

. The fact that competitors haven't uniformly adopted similar restrictions suggests the industry remains divided on how to balance capability with legal exposure. Writers and users should watch for potential regulatory developments that could standardize these practices across platforms, as lawmakers consider rewriting AI copyright rules in response to ongoing litigation 4

4

.