July 28, 2026. A federal judge dismissed most of Google's DMCA lawsuit against web scraper SerpApi. The ruling exposes a company that built a trillion-dollar empire scraping the web now trying to criminalize the same practice. Here is what the court actually held, what comes next, and why the stakes go far beyond one Austin startup.
Status: Google v. SerpApi partially dismissed with prejudice (July 20, 2026). Google has until August 10 to amend. Reddit v. SerpApi, Perplexity, Oxylabs, AWMProxy oral argument held July 23; motions pending.
Two federal lawsuits, filed within two months of each other. Both aimed at SerpApi, an Austin-based company that scrapes Google search results and sells structured access through an API. Both built on the same unusual legal theory. Both now facing the same headwinds.
Google v. SerpApi (N.D. Cal., No. 4:25-cv-10826, Judge Yvonne Gonzalez Rogers). Filed December 19, 2025. Google claimed SerpApi violated the Digital Millennium Copyright Act by circumventing SearchGuard, Google's anti-bot system. The complaint alleged browser spoofing, IP rotation, and automated challenge-solving.
Reddit v. SerpApi, Perplexity, Oxylabs, AWMProxy (S.D.N.Y., No. 1:25-cv-08736, Judge Paul Engelmayer). Filed October 22, 2025. Reddit accused the defendants of circumventing both Reddit's own anti-scraping measures and Google's SearchGuard to harvest Reddit content through Google search results. The complaint alleged that in two weeks in July 2025, the defendants scraped nearly three billion Google SERPs containing Reddit material. Reddit's filing compared the defendants to would-be bank robbers who, unable to enter the vault, break into the armored truck instead.
Reddit also deployed a forensic test: it created a post containing a unique hexadecimal string served only to Google's crawler. That string later appeared in a Perplexity response. The evidence that Reddit data flows through Google search results into AI systems is strong. The legal question is whether the DMCA is the right tool to stop it.
The July 20 ruling was a split decision.
Dismissed permanently: Claims based on ordinary search results containing no copyrighted content. URLs, factual snippets, and uncopyrightable data cannot support a DMCA Section 1201 claim. The DMCA's anti-circumvention provisions apply only when a technological measure controls access to a work protected by copyright. This part of Google's case is dead.
Dismissed temporarily: Claims involving copyrighted components in Knowledge Panels. Google has until August 10 to fix a critical deficiency: it never alleged that the actual copyright holders authorized SearchGuard to protect their works. Section 1201(a)(3)(B) of the DMCA requires that a technological measure operate with the authority of the copyright owner. Google simply asserted that authority without evidence.
The court rejected two of SerpApi's arguments: that only copyright owners can sue under the DMCA, and that browser spoofing cannot constitute circumvention as a matter of law. These are procedural losses for SerpApi. But Google lost the case.
SerpApi's CEO Julien Khaleghy calculated the damages using Google's own numbers. Even at the low end of the DMCA's statutory range of $200 to $2,500 per violation, applied to the query volume Google alleged, the damages would reach $7.06 trillion. That exceeds the entire GDP of the United States.
When the logical endpoint of a legal theory produces damages larger than the American economy, the theory is the problem. Congress did not design the DMCA to turn every automated web request into a federal copyright violation carrying seven trillion dollars in statutory exposure.
Google says it is protecting copyright holders. But SerpApi's motion to dismiss surfaced something Google's own complaint admitted: SearchGuard exists to protect Google's advertising business. The DMCA requires a technological measure to control access to a copyrighted work. It does not protect revenue streams.
Reddit says it is protecting its users. But Reddit's user agreement states that users retain copyright in their posts. Reddit does not own the content it is suing over. It holds a license. What Reddit owns is a content licensing deal with Google reportedly worth sixty million dollars annually. The scrapers were getting for free what Google was paying for. This is about protecting a revenue line, not users.
As Meredith Rose, senior policy counsel at Public Knowledge, told Ars Technica: Google and Reddit are grasping at whatever tool is available in the face of accelerating AI scraping. The DMCA was not designed for this. But historically, it has been an effective way to force licensing conversations.
Google's only path forward is to amend its complaint to focus on Knowledge Panel content it licenses from third parties. To revive its case, Google must argue that Knowledge Panels are full of copyrighted material and that SearchGuard was deployed with the copyright holders' authority to protect it.
Rose identified the danger in this argument: if Google claims Knowledge Panels contain copyrighted works, it invites the question of whether Google licensed all of them. If Knowledge Panels contain works displayed without a license, Google has admitted to infringement on a massive scale. If Google did license everything, it must produce those licenses and show each rights holder authorized SearchGuard. Either way, Google faces exposure.
On July 23, Judge Engelmayer heard nearly three hours of oral argument in Manhattan. According to Bloomberg Law, the judge showed clear skepticism about Reddit's authority to sue over copyrights held by its users. The motions remain pending, but the questioning suggests the same standing problem that sank most of Google's case.
The California ruling is not binding on New York. But the statutory interpretation question is identical: does the DMCA's anti-circumvention provision apply to an anti-bot system that does not distinguish between copyrighted and uncopyrighted content, deployed without the authority of copyright owners? Two federal judges now appear to be converging on the same answer.
Rose calls what is happening the re-enclosure of the web. Since 2023, platforms have been cutting off access that was previously open, using whatever legal tools they can find. The victims are not just AI companies. They include researchers studying online discourse, journalists investigating platform behavior, archivists preserving digital history, and public health officials tracking misinformation.
The DMCA is a uniquely dangerous tool for this purpose because Section 1201 makes circumvention illegal regardless of whether the underlying use is lawful. You can circumvent a technological protection measure for a perfectly legal purpose and still violate the statute. It is a strict-liability regime dressed up as copyright law.
As Mike Masnick wrote at Techdirt in December 2025, days after Google filed: Google built its empire scraping the web. Now it is suing to stop others from scraping Google. The company whose original mission was to organize the world's information and make it universally accessible is now arguing in federal court that accessing public web pages without permission is a federal crime.
August 10, 2026: Google's deadline to amend. The company says it looks forward to filing. But the path is narrow. Google must thread a needle that may not exist: claim enough copyright to satisfy the DMCA without admitting so much that it exposes itself to infringement liability.
Reddit's ruling is expected within weeks. Judge Engelmayer's skepticism at oral argument, combined with the Google ruling, suggests Reddit faces similar headwinds.
The broader stakes: if Google and Reddit succeed, every major platform gets a DMCA cudgel to demand licensing fees from anyone who wants to access public data at scale. The internet's founding bargain was that anything on a public web page is accessible to anyone with a browser. Google and Reddit are asking the courts to rewrite that bargain retroactively. Judge Rogers just told them no.
Whether Judge Engelmayer agrees, and whether Google's Hail Mary on Knowledge Panels survives its own contradictions, will determine whether the DMCA becomes a gatekeeper for the open web or remains what Congress intended: a tool for stopping piracy, not for controlling access to public information.
This analysis is based on publicly available court filings and does not constitute legal advice. For questions about DMCA claims, consult an attorney.