What Does the New York Times' 'Dual Strategy' Say About the AI Era?

In late May 2025, The New York Times signed an AI content license agreement with Amazon, permitting Amazon's AI-based services (Alexa) and generative AI model training to formally use NYT news articles, cooking content (NYT Cooking), and sports content (The Athletic). The contract is multi-year; Amazon became a leader in the "formal data licensing model" — paying rights holders rather than training on crawled content. Background: in late 2023, NYT (with NY Daily News, 8 regional newspapers, Center for Investigative Research Inc.) sued OpenAI and Microsoft for copyright infringement — claiming unauthorized content collection and AI model training use, with ChatGPT reproducing NYT articles nearly verbatim, constituting commercial infringement beyond fair use. On April 4, 2025, US SDNY Judge Sidney Stein dismissed DMCA violation and unfair competition claims but allowed the AI model training data use claim and AI output copyright infringement possibility to proceed. Current status: NYT gathering evidence for trial. The dual strategy: suing OpenAI while licensing to Amazon is not contradictory but strategically parallel — "unauthorized use is illegal, authorized use with fair compensation is acceptable." NYT's position articulates: news organizations should be active participants in determining how AI uses their content, not passive victims; licensing agreements establish precedent that news content has monetizable value for AI training; litigation against unauthorized use maintains the leverage necessary to command meaningful licensing fees. Industry significance: the NYT precedent is creating a bifurcated AI training data market — companies willing to pay (Amazon, Apple, others) securing agreements while holdouts face litigation.