Britannica and Merriam-Webster File Lawsuit Against OpenAI Over AI ‘Memorization’ Claims

Encyclopedia Britannica and Merriam-Webster have launched a legal challenge against OpenAI alleging unauthorized use of copyrighted material in training its AI systems. The publishers claim OpenAI’s ChatGPT reproduced content “substantially similar” to their encyclopedias and dictionaries after allegedly copying their works without permission. Britannica asserts that GPT-4 has effectively “memorized” substantial portions of its proprietary content creating near-verbatim outputs that infringe on their intellectual property rights. The lawsuit filed Friday seeks damages and an injunction to prevent further alleged misuse of their copyrighted material in AI training datasets. This case represents a pivotal moment in the evolving conflict between traditional content creators and AI developers over fair use and compensation in the digital age. Legal experts suggest the outcome could establish critical precedents for how AI companies must handle copyrighted works during model training. OpenAI has not yet issued a formal response to the allegations but faces mounting pressure from publishers and creators worldwide demanding compensation for content used to develop generative AI systems. The case highlights growing tensions as AI capabilities expand and questions of intellectual property ownership become increasingly complex in the machine learning era.

Dispatch from Zara Blackwood (Rapid Response Intelligence Analyst).