A Manhattan federal judge has allowed Reddit to proceed with core DMCA claims against SerpApi and Perplexity AI, despite the defendants’ efforts to dismiss the case. The lawsuit alleges the companies circumvented Google’s anti-bot protections to scrape Reddit content at scale.
The legal battle centers on whether a platform can claim protection under the Digital Millennium Copyright Act (DMCA) when the “lock” being broken isn’t their own. In this instance, Reddit is suing over the alleged bypass of security measures implemented by Google, not its own internal systems.
The Standing Dispute: Who Owns the Lock?
The case is viewed by some legal observers as unconventional. Meredith Rose, a DMCA expert with Public Knowledge, noted that the legal standing for such a claim is typically reserved for the copyright owner, an exclusive licensee, or the entity that actually deployed the technological protection measure. According to Rose, Reddit is none of those things
in this scenario.
Despite these hurdles, the court has decided to hear the facts of the case. The judge did, however, toss Reddit’s claims regarding unfair competition and unjust enrichment, ruling that these were preempted by the Copyright Act.
SerpApi’s Defense and the Nature of Public Data
SerpApi and Perplexity AI argue that they are not attacking Reddit’s platform, but are instead accessing information already available via public search results. Jeff Homrig, a lawyer for SerpApi, contends that the facts of the case favor the defendants, asserting that public information does not become protected simply because a platform intends to charge for it.
“we remain confident in our position. The court has decided to hear the facts; the facts are on our side. SerpApi accesses public search results, not Reddit’s platform, and public information does not become protected because a platform wants to charge for it. We look forward to making that case.”
Jeff Homrig, a lawyer for SerpApi, via Ars Technica
The outcome may hinge on discovery. One possibility is that SerpApi and Perplexity AI could prove Reddit never actually authorized Google to protect its content in search results. Additionally, if the defendants can establish that publicly accessible content in Google search results is not protected by the Copyright Act, their defense would be significantly strengthened.
The $60 Million Tension with Google
This legal friction coincides with a broader breakdown in the relationship between Google and the publishers it relies on. Reddit and Google are currently negotiating a renewal of a licensing agreement signed in 2024, which reportedly pays Reddit around $60 million per year.
While the deal provided Reddit with significant capital ahead of its IPO and gave Google a steady stream of human conversations for AI training, the value proposition is shifting. Google’s AI products now frequently use this data to answer questions directly in search results, meaning users often find the answer without ever clicking through to Reddit.
The financial stakes are high enough to rattle the market; Reddit shares fell as much as 5.8% in premarket trading following reports that the company had discussed blocking Google from using its user posts for AI purposes.
The Collapse of Referral Traffic
Reddit is not alone in its frustration. Several major publishers are reconsidering their access to Google as organic referral traffic plummets. Data from Semrush highlights a steep decline in U.S.
| Publisher | Reported Traffic Decline |
|---|---|
| Business Insider | More than 85% |
| CNN | Around 25% |
| Politico | 23% |
| USA Today | Almost half |
Mike Reed, USA Today Co. CEO, expressed the sentiment of many publishers, stating, It’s time to take a stand and say enough is enough
.
The dilemma for these publishers is a “catch-22” of visibility. While they can use Google-Extended to stop Gemini from training on their data, preventing content from appearing in AI Overviews risks removing them from conventional search results entirely. This is a critical distinction: blocking the AI often means blocking the visibility that drives the business.
Regulatory Interventions and the AI Transition
The UK is attempting to break this deadlock. The Competition and Markets Authority recently ordered Google to allow publishers to opt out of AI search features without losing their place in traditional search results. Google has stated it will test these controls in Britain before a potential global rollout, though no specific timeline has been provided.
The urgency for these changes is driven by a stark shift in user behavior. A Pew Research Center study found that in searches without an AI Overview, users clicked traditional links in 15% of cases. When an AI Overview appeared, that number dropped to 8%, and links within the summaries themselves were clicked in only 1% of searches.
As the DMCA case moves forward, the core question remains whether the law can adapt to a world where “protection” is managed by a third-party intermediary like Google, and whether platforms can successfully sue others for bypassing a lock they didn’t actually build.
Worth a look
