Reddit Inc. vs. SerpApi LLC (S.D.N.Y. 1:25-CV-08736) [pdf] (storage.courtlistener.com)

Reddit sued SerpApi, Oxylabs, AWMProxy and Perplexity in the Southern District of New York, accusing them of “industrial-scale” circumvention of Reddit’s access controls to harvest copyrighted user content. The complaint alleges the scraper vendors and proxy services evaded Reddit’s defenses by scraping Reddit material from Google search engine results pages (SERPs) — not Reddit’s site directly — and that during a two-week window in July 2025 the defendants automated access to nearly three billion SERPs containing Reddit text, URLs, images and videos. Reddit also says Perplexity, an “answer engine” and customer of SerpApi, continued to ingest Reddit content after a cease-and-desist (its Reddit citations rose ~40x) and that at least one vendor (AWMProxy) is tied to a previously criminal botnet. The case is significant for AI/ML because it targets the data supply chain many models rely on: vendor-enabled circumvention (proxy networks, residential IPs, location spoofing, human-mimicking scrapers) and SERP-based harvesting. Reddit asserts DMCA anti-circumvention claims (17 U.S.C. §1201), seeking injunctions and damages — a ruling could broaden liability for scraper-as-a-service providers and their AI customers, push more companies toward licensing deals (as OpenAI and Google have done with Reddit), and force practitioners to re-evaluate data provenance, vendor due diligence, and downstream legal exposure when training or fine-tuning models on web-derived content.