CNN says Perplexity is not merely learning from CNN in some abstract model-training sense, but repeatedly copying CNN content, retrieving it in real time,...
Summary
CNN has filed a lawsuit against Perplexity, alleging the AI "answer engine" built its commercial product by systematically copying, scraping, summarizing, and reproducing CNN journalism without authorization or payment. The complaint, which implicates more than 17,000 CNN works, focuses on Perplexity's real-time content retrieval and output generation, rather than abstract model training. Key allegations include verbatim or near-verbatim reproductions, the bypass of CNN's paywalls to deliver full articles to non-subscribers, failed licensing negotiations in October 2025, and Perplexity's continued use of content after a December 2025 cease-and-desist letter. This case is significant as it shifts AI copyright litigation from training data disputes to the more concrete issue of whether AI tools are directly replacing the market for original journalism.
Key takeaway
For legal professionals advising publishers on AI strategy, this lawsuit highlights the critical need to document direct content reproduction and market substitution. You should focus on evidence of verbatim outputs, paywall circumvention, and failed licensing negotiations, as these strengthen claims of willful infringement. Proactively test AI services for unauthorized use of your content across all product tiers, including APIs, and ensure clear cease-and-desist letters are issued to establish notice.
Key insights
AI copyright litigation is shifting from training data to direct content reproduction and market substitution by "answer engines."
Principles
- Verbatim output and paywall bypass strengthen copyright infringement claims.
- Failed licensing talks demonstrate intent and knowledge of unauthorized use.
- AI-assisted distribution and substitution are new litigation focal points.
Method
Publishers can build AI copyright cases by collecting outputs, testing paid/free versions, documenting blocking, and preserving licensing negotiation records.
In practice
- Test AI products for verbatim reproduction of your paywalled content.
- Document all attempts to block crawlers and negotiate licenses.
- Frame market harm around lost subscriptions, advertising, and licensing.
Topics
- AI Answer Engines
- Copyright Litigation
- Content Licensing
- Perplexity AI
- Journalism Economics
- Paywall Bypass
Best for: CTO, VP of Engineering/Data, Investor, Legal Professional, Tech Journalist, Director of AI/ML
Related on AIssential
Editorial summary, takeaway, and curation by AIssential. Original article published by Pascal’s Substack.