OpenAI Limits ChatGPT Author

OpenAI has introduced new limits on ChatGPT that block prompts asking the system to write in the style of a named author. Under the updated behavior, the model is designed to refuse direct imitation of a specific writer’s voice and instead steer users toward outputs built around broader characteristics, such as genre, tone, or thematic elements.

The change applies regardless of whether the referenced author is living or deceased, according to the description of the restriction. That marks a shift from earlier model behavior, which had been more willing to comply with requests to mimic particular authorial styles.

How ChatGPT responses are changing

When a user requests text that explicitly copies When a user requests text that explicitly copies a named writer’s style, the system now declines and offers an alternative approach. The alternative is positioned as a way to capture general qualities—such as pacing, mood, or narrative structure—without anchoring the output to a specific person’s identifiable voice. This design aims to draw a clearer boundary between creative inspiration and direct stylistic imitation. The stated intent is not to stop users from generating similar kinds of writing, but to avoid outputs framed as replicas of a particular author. Copyright lawsuits and legal risk in the US The policy adjustment is described as appearing to be a strategic response to ongoing copyright litigation involving the company. Plaintiffs in current lawsuits have argued that the model’s capability to reproduce copyrighted stylistic expression amounts to infringement.

Officials involved in these disputes have focused on Officials involved in these disputes have focused on the legal concept that, even when “style” is not itself protected, outputs may still pose risk if they are considered too close to protected expression. The concern highlighted is “substantial similarity,” a standard that can become central when deciding whether a work has crossed from general inspiration into unlawful copying.

In that context, limiting direct mimicry is presented as a way to reduce exposure to claims that the system is enabling unauthorized reproduction of protected creative works.

Pressure from writing organizations and uneven industry rules

The shift also aligns with wider industry pressure from professional writing organizations, which have raised concerns that AI-generated writing can create unfair competition. Those groups have pushed for guardrails that prevent tools from generating output that competes by echoing recognizable creative identities too closely.

At the same time, policies across major AI developers remain inconsistent. Some models continue to comply with style-mimicry requests, while others have adopted measures similar to the restrictions described here.

For users, the immediate practical impact is a narrower set of instructions that will be accepted, with more emphasis on describing desired qualities instead of naming a specific author. For developers and rights holders, the change underscores how legal disputes and industry pressure are shaping product behavior in real time, even as the boundaries of protection for style versus expression remain contested in court.

Implications

Country Impact: In the United States, the move is tied to ongoing copyright litigation and centers on how courts treat “substantial similarity” claims. The change may influence how users frame requests by shifting away from named-author prompts toward general descriptive instructions.

Industry Impact: For publishing and professional writing groups, the restriction reflects growing pressure to curb AI-enabled imitation that may be seen as unfair competition. For AI developers, it highlights a compliance-oriented approach that can translate legal and reputational concerns into product-level guardrails.

Market Impact: The update signals that competitive differentiation may increasingly include how strictly models handle author-style prompts. Uneven policies across major developers could shape user expectations and platform adoption depending on how permissive or restrictive different systems remain.