A Manhattan federal judge has allowed Reddit's copyright lawsuit against Perplexity AI and SerpApi to proceed, rejecting bids to dismiss core DMCA claims. Judge Paul A. Engelmayer ruled that Reddit has standing to sue over the scraping of user-generated content due to its unique arrangement of posts, and that bypassing technical barriers like Google's SearchGuard constitutes actionable reputational harm. While state-law claims were dismissed, the decision provides significant legal leverage to publishers defending their content against unauthorized AI training crawlers.
Federal copyright lawsuit ruling
- ▪Judge Paul A. Engelmayer dismissed Reddit Inc.'s state-law claims for unjust enrichment and unfair competition against Perplexity AI Inc. and SerpApi LLC.
- ▪Reddit Inc. alleges that Perplexity AI Inc. and SerpApi LLC bypassed technological protections, including Google's SearchGuard, to scrape user content at scale
- ▪U.S. District Judge Paul A. Engelmayer rejected a motion by Perplexity AI Inc. and SerpApi LLC to dismiss the core of Reddit Inc.'s copyright lawsuit
User-generated content copyright standing
- ▪Perplexity AI Inc. and SerpApi LLC argued that Reddit Inc. lacked standing to bring the lawsuit because the platform's content is created by its users
- ▪The U.S. District Court for the Southern District of New York ruled that Reddit Inc. has independent copyright standing to sue over user-generated content due to its protected arrangement of that content
Technical circumvention reputational harm
- ▪The Digital Millennium Copyright Act allows civil actions by any injured person, which the court ruled applies to Reddit Inc.'s claims of technical circumvention
- ▪The federal court ruled that circumventing a platform's technical protections constitutes reputational harm sufficient to sustain a lawsuit, even without demonstrating direct economic loss
AI crawler traffic impact
- ▪Reddit Inc. has established licensing agreements for AI training with companies including Google and OpenAI, which it argues are undermined by unauthorized scraping
- ▪AI search engines ingest publisher content at scale and return synthesized answers to users while sending minimal referral traffic back to the original sources
Publisher anti-scraping protections
- ▪Publishers can implement explicit disallow directives for known AI crawler user agents within their robots.txt files to document non-consent to scraping
- ▪The court's recognition of technical circumvention as harm validates the legal standing of publishers who implement anti-scraping measures and robots.txt protocols
Story comments
Loading comments…