Google v. SerpApi: Federal Court Rejects Google's Broad DMCA Anti-Scraping Theory — What the Decision Means for AI Companies, Search Tools, Developers, and Website Owners
Google Tried to Turn the DMCA Into an Anti-Scraping Law. A Federal Judge Said "Not So Fast."
In one of the most closely watched internet law decisions of 2026, Chief U.S. District Judge Yvonne Gonzalez Rogers of the Northern District of California significantly limited Google's attempt to use the Digital Millennium Copyright Act ("DMCA") as a weapon against web scraping. The case, Google LLC v. SerpApi, has implications that extend well beyond Google and one API provider. It touches virtually every business that collects publicly available internet data, every AI company training on public information, every SEO platform, every price comparison service, every researcher, and every creator whose copyrighted content appears online.
For years, courts have wrestled with the legal boundaries between protecting intellectual property and preserving an open internet. Cases involving the Computer Fraud and Abuse Act ("CFAA"), copyright infringement, terms of service, and privacy laws have all shaped the rules governing web scraping. Google attempted to add another powerful weapon to its arsenal by arguing that bypassing its anti-bot technology violated the DMCA's anti-circumvention provisions, even where the information being accessed was publicly available.
Judge Gonzalez Rogers rejected much of that theory. Although Google was given an opportunity to amend part of its complaint relating to copyrighted images appearing in Knowledge Panels, the court permanently dismissed Google's broader argument that scraping ordinary search results constitutes copyright circumvention under the DMCA. The decision is likely to become one of the leading authorities interpreting Section 1201 of the DMCA in the context of web scraping and artificial intelligence.
Background of the Lawsuit
Google filed suit against SerpApi in late 2025, alleging that SerpApi operated an API service allowing customers to retrieve Google search results by automatically sending queries to Google's search engine. According to Google's complaint, SerpApi allegedly bypassed Google's anti-bot technology, known as SearchGuard, which Google claimed was designed to prevent automated scraping of Google Search.
Rather than asserting a traditional copyright infringement claim, Google pursued a more aggressive legal theory under 17 U.S.C. § 1201, the anti-circumvention provisions of the DMCA.
Section 1201 generally prohibits circumventing technological measures that effectively control access to copyrighted works. Congress originally enacted these provisions to combat digital piracy, including the circumvention of encryption protecting DVDs, software, digital music, and other copyrighted media.
Google argued that SearchGuard constituted such a technological protection measure and that SerpApi violated federal law by bypassing it.
The Legal Issue Before the Court
The central question presented to the court was remarkably straightforward:
Can a company use the DMCA's anti-circumvention provisions to prevent scraping of publicly available search results simply because it has implemented anti-bot technology?
Stated another way:
Does a website operator transform publicly available information into protected copyrighted content merely by placing anti-bot software in front of it?
Judge Gonzalez Rogers concluded that, for most search results, the answer is no.
Understanding Section 1201 of the DMCA
The DMCA is often misunderstood.
Most people associate copyright law with copying protected works without permission.
Section 1201 is different.
Instead of prohibiting copying itself, Section 1201 prohibits circumventing technological measures that effectively control access to copyrighted works.
Classic examples include:
-
breaking DVD encryption;
-
bypassing software activation systems;
-
defeating password protection protecting copyrighted files;
-
decrypting digital media.
Congress intended these provisions to prevent digital piracy—not to create a general prohibition against automated access to publicly available information.
That distinction became the centerpiece of the court's analysis.
The Court Divided Google's Claims into Two Separate Categories
Rather than treating all search results identically, Judge Gonzalez Rogers carefully distinguished between two different situations.
Category One: Ordinary Search Results Without Copyrighted Content
The first category involved ordinary Google search results consisting primarily of:
-
hyperlinks;
-
factual information;
-
website titles;
-
snippets;
-
publicly available information.
The court held that these search results generally do not constitute copyrighted works protected by Section 1201.
Because Section 1201 protects technological measures controlling access to copyrighted works, the statute simply does not apply where no copyrighted work is involved.
Accordingly, the court dismissed these claims with prejudice, meaning Google cannot reassert this theory in this case.
This portion of the opinion represents perhaps the most significant aspect of the decision because it establishes a clear legal principle:
No copyrighted work means no DMCA anti-circumvention claim.
Category Two: Search Results Containing Copyrighted Components
The second category involved search results containing copyrighted material, particularly licensed photographs displayed in Google's Knowledge Panels.
Unlike ordinary search snippets, these images clearly constitute copyrighted works.
However, the court found another fatal defect in Google's complaint.
Section 1201 requires that the technological protection measure operate with the authority of the copyright owner.
Google did not allege facts showing that the owners of those licensed images authorized Google to deploy SearchGuard on their behalf.
Because that statutory element was missing, the court dismissed these claims as well.
Unlike the first category, however, these claims were dismissed without prejudice, allowing Google twenty-one days to amend its complaint.
Why Google's Theory Failed
The opinion repeatedly emphasizes that SearchGuard functions primarily as an anti-bot mechanism rather than as a technology designed to protect copyrighted works.
The court observed that Google's own allegations described SearchGuard as regulating automated access to search results generally.
That distinction proved critical.
An anti-bot system protecting an entire website is fundamentally different from a technological measure protecting access to a particular copyrighted work.
The DMCA does not convert every website access restriction into copyright protection.
Instead, the statute requires a direct nexus between the technological measure and the copyrighted work itself.
Without that connection, Section 1201 does not apply.
What the Court Did NOT Decide
It is equally important to understand what the court did not hold.
The decision does not declare web scraping universally lawful.
Numerous legal theories remain available depending upon the facts, including:
-
breach of contract based upon Terms of Service;
-
traditional copyright infringement;
-
Computer Fraud and Abuse Act claims in appropriate jurisdictions;
-
state unfair competition laws;
-
privacy statutes;
-
trade secret claims;
-
misappropriation doctrines.
The opinion addresses only Google's attempt to invoke the DMCA's anti-circumvention provisions.
Businesses should not interpret the decision as a blanket authorization to scrape every website on the internet.
The Relationship to AI and Machine Learning
Although the lawsuit concerns Google Search, its implications extend directly into artificial intelligence.
Modern AI systems routinely collect publicly available information to train language models, build search indexes, perform retrieval-augmented generation, conduct market research, and analyze public data.
Website owners increasingly seek legal mechanisms to restrict those activities.
This opinion suggests that merely deploying anti-bot technology does not automatically convert publicly available information into content protected by Section 1201.
That principle may become increasingly important as courts confront disputes involving AI training datasets and automated collection of public information.
The Connection to hiQ Labs v. LinkedIn
The decision also fits within a broader line of cases protecting access to publicly available internet information.
In hiQ Labs, Inc. v. LinkedIn Corp., the Ninth Circuit held that scraping publicly accessible LinkedIn profiles generally does not violate the Computer Fraud and Abuse Act because public websites are not protected computers requiring authorization in the same manner as private systems.
Google attempted to achieve under the DMCA what LinkedIn unsuccessfully sought under the CFAA.
Judge Gonzalez Rogers declined to expand Section 1201 into a comparable anti-scraping statute.
Together, hiQ and Google v. SerpApi reinforce an important legal distinction between protecting genuinely restricted information and attempting to monopolize access to public internet data.
Practical Takeaways for Businesses
Businesses operating websites should carefully review this decision before assuming anti-bot technology alone provides protection under the DMCA.
Companies should consider:
-
strengthening contractual Terms of Service;
-
implementing appropriate licensing agreements;
-
clearly identifying copyrighted content;
-
distinguishing between factual data and creative works;
-
developing layered legal strategies rather than relying solely upon the DMCA.
The opinion illustrates that copyright law has limits, particularly where publicly available factual information is concerned.
Practical Takeaways for AI Companies
Artificial intelligence companies should not view this decision as a free pass.
However, the opinion provides meaningful guidance.
AI developers should continue:
-
avoiding unauthorized copying of copyrighted creative works;
-
respecting robots.txt files where appropriate;
-
complying with contractual obligations;
-
minimizing collection of copyrighted images and expressive works when unnecessary;
-
carefully documenting lawful data collection practices.
The court drew an important distinction between collecting public factual information and reproducing copyrighted creative content.
Practical Takeaways for Content Creators
Photographers, artists, journalists, and creators should also understand this decision.
The opinion does not diminish traditional copyright protections.
Creators still retain exclusive rights under the Copyright Act.
Indeed, Google's remaining claims involving licensed Knowledge Panel images demonstrate that copyrighted creative works continue to receive legal protection under appropriate circumstances.
The court simply required Google to satisfy the statutory requirements before invoking the DMCA.
Practical Takeaways for SEO Companies and Data Aggregators
Businesses engaged in SEO analysis, market intelligence, competitive research, or price comparison should pay close attention.
The decision suggests courts may distinguish between:
-
collecting publicly available factual information; and
-
copying copyrighted expressive content.
That distinction may significantly reduce DMCA exposure while leaving other legal risks intact.
Companies should nevertheless maintain robust compliance programs addressing copyright, privacy, contractual obligations, and applicable state and federal laws.
Conclusion
Google v. SerpApi represents one of the most significant internet law decisions of 2026. Judge Gonzalez Rogers declined to expand the DMCA into a broad anti-scraping statute, holding that Section 1201 requires a technological protection measure controlling access to an actual copyrighted work—not merely an anti-bot system protecting publicly available information. While Google may continue pursuing a narrower claim involving licensed Knowledge Panel images, the court permanently rejected its broader theory that bypassing SearchGuard to access ordinary public search results constitutes copyright circumvention.
For businesses, developers, AI companies, creators, and website operators, the decision reinforces an important principle: copyright law protects copyrighted works—not access restrictions standing alone. At the same time, companies should remember that web scraping remains subject to a variety of other legal doctrines, including contract law, copyright infringement, privacy statutes, and, in some cases, the Computer Fraud and Abuse Act. As courts continue to define the legal boundaries of AI and automated data collection, Google v. SerpApi is likely to become a foundational case in the evolving law governing public internet information.
Need Advice on Copyright, Web Scraping, AI, or DMCA Issues?
Vondran Legal® represents businesses, software developers, AI companies, content creators, photographers, technology startups, and online platforms in complex copyright, DMCA, software licensing, and internet law matters nationwide. If you have received a DMCA notice, are involved in a web scraping dispute, or need guidance regarding copyright compliance or AI-related legal risks, contact Vondran Legal to schedule a confidential consultation and discuss your options.
