When the oldest authority of knowledge meets the most radical AI giant, a copyright battle that could reshape industry rules has begun in the US District Court in Manhattan.

Recently, Encyclopaedia Britannica and its subsidiary Merriam-Webster have formally sued OpenAI. The core of this lawsuit targets the "original sin" of generative AI: whether using nearly 100,000 high-quality encyclopedic articles without authorization to train models is an "acceptable use" of technological innovation or a "systematic infringement" of human intellectual heritage?

Debate over Input: Is it "transformative use" or "riding on the coattails"?

In the complaint, Encyclopaedia Britannica accused ChatGPT of "riding on the coattails." They believe that encyclopedia entries are original copyrighted products carefully compiled, with high commercial value. OpenAI, however, argues that their training is based on publicly available data and complies with the "fair use" principle, representing a transformative act that promotes innovation.

Evidence of Output: Does AI "memorize" the original text?

The most damaging evidence in the lawsuit lies in the "memorized" reproduction. Encyclopaedia Britannica submitted comparisons showing that under specific prompts, ChatGPT produces content that is "identical or highly similar" to the original. This means AI is not only learning knowledge but also directly distributing copyrighted original texts to users without payment, even "siphoning" traffic from the original platform through generated summaries.

Brand Crisis: When AI's "hallucinations" are attributed to authoritative names

Aside from copyright, this lawsuit also involves trademarks and attribution rights. Encyclopaedia Britannica accuses OpenAI of improperly citing its brand in AI-generated misinformation. For an authoritative institution with over 250 years of history, if AI's "serious nonsense" is attributed to its name, it would damage the credibility of knowledge on which it relies.

Industry Trendsetter: A Battle for Source Order

Currently, Encyclopaedia Britannica has requested the court to issue an order to stop the infringement and claim damages. Industry experts analyze that this case may be merged with the New York Times vs. OpenAI case for trial.

As startups like Perplexity AI also fall into similar disputes, traditional knowledge institutions are collectively waging war against the "freeloaders" of the AI era. The outcome of this verdict will determine the cost bottom line for future AI development and whether the heights of human original knowledge can still maintain their due dignity in the age of rapid computing power.