Google has opened a fresh legal front in a fight that’s quietly reshaping how the internet’s plumbing gets used: a federal lawsuit against SerpApi, a small Texas company that sells a straightforward pitch to developers — an unofficial, paid way to pull structured Google search results. In its complaint, filed Dec. 19, 2025, Google says SerpApi didn’t just scrape a few pages for convenience; it “siphoned” results at scale, running hundreds of millions (Google’s word) of automated queries while cloaking those requests to look like ordinary human searches.

At the heart of Google’s case is a security layer the company calls SearchGuard, which Google says it rolled out in January 2025 to spot and filter automated scraping of search results. The complaint alleges SerpApi built a toolkit to evade those protections: fake browsers, rotating IPs, spoofed device and location signals — all the tricks a sophisticated crawler can use to blend into human traffic. In Google’s telling, that wasn’t curiosity or experimentation; it was a deliberate plan to turn Google Search into a backdoor API and then commercialize the output.

Why would Google care so much about search results? Because modern search pages are no longer just lists of links. They’re composite products: licensed images, aggregated facts, maps, shopping panels and knowledge boxes — some content Google licenses from partners or pays third parties to display. Google argues that when scrapers extract those results wholesale, they’re not only leeching off Google’s engineering but also repackaging other people’s copyrighted material that appears inside the SERP (search engine results page). The company has framed the conduct not merely as a contract breach or abuse of access, but as an anti-circumvention problem under the DMCA — meaning each bypass of SearchGuard could expose a defendant to statutory damages.

That legal framing is important because it raises the stakes. Anti-circumvention claims under the DMCA let plaintiffs pursue per-violation statutory damages that can balloon quickly — a natural deterrent to this kind of scraping if courts buy Google’s account. Google is asking for an injunction to stop the alleged circumvention, for any circumvention tools to be destroyed, and for damages that could in theory be substantial. But the company also notes the arithmetic: SerpApi publicly reports only a few million dollars in annual revenue, which complicates how much any judgment could realistically recover.

SerpApi’s product is simple and, to many developers, useful: since Google doesn’t publish a public API for search results, SerpApi provides a RESTful interface that returns parsed, structured SERPs for customers who want to build SEO dashboards, research tools, or feed other systems. From that perspective, the company is filling a void — and when the product hits a nerve, it scales quickly. SerpApi has pushed back publicly, saying it provides what’s visible in ordinary browsers and implying that its business model rests on publicly accessible pages rather than secret access. The company has already faced litigation dustups: Reddit sued SerpApi (and others) in October 2025 for allegedly siphoning Reddit content via Google as part of data pipelines used by AI firms.

Those earlier fights offer context. Platforms and publishers have grown wary of large-scale scraping because scraped web content has been an ingredient for many AI models and products — sometimes used without licenses or permission. Reddit’s suit accused scrapers of using Google as a conduit to reach content that Reddit had tried to restrict; in a now-notorious maneuver, Reddit created a “trap” post that was only discoverable by crawlers, then found that content resurfaced in third-party AI outputs. That episode demonstrated how easily third parties can combine scraping with clever technical workarounds to access content that rightsholders assumed was shielded.

There’s an irony in Google’s posture. For decades, Google’s own bread and butter has been crawling and indexing vast swaths of the open web — a business model that critics have long argued allowed Google to monetize other people’s content. Now, with search pages themselves becoming product inputs for AI and services, Google is trying to fence off parts of the experience it considers either copyrighted or licensed, and to use copyright and anti-circumvention law to do it. Whether the courts will accept Google’s novel application of the DMCA will matter not just for SerpApi but for any company that turns user-facing web pages into machine-friendly APIs.

The case also highlights a tensions map that runs through the modern web: publishers who want control and licensing fees, platforms who want to protect partners and their own assets, startups who see value in making web output machine-readable, and AI firms that want as much high-quality public data as they can legally get. Google’s complaint argues that SerpApi’s resale of structured results undermines licensing deals and makes licensed imagery available to rivals who otherwise wouldn’t pay. Defenders of open access counter that a lack of a public API leaves a product gap that market players inevitably try to fill. Expect lawyers on both sides to parse the fine line between “publicly viewable” and “lawfully extractable.”

What are the possible outcomes? If Google wins a permanent injunction and the court endorses its anti-circumvention theory, the decision could chill an entire slice of tooling around SERPs: SEO scrapers, third-party dashboards, and any firm that treats search pages as an official data source. If the court rejects the DMCA angle or finds the technical facts insufficient to show effective circumvention, scrapers will still face contract and trespass-to-chattels claims, but the decision would leave more breathing room for companies that provide machine interfaces to human-facing pages. Either way, appeals are likely, and the case — like many modern platform fights — may be argued before appellate courts and could take years to reach a definitive answer.

There are knock-on effects for AI development, too. Many startups and research groups rely on what’s technically feasible, and if the legal environment suddenly makes certain collections or pipelines hazardous, product roadmaps and training data sources will need rethinking. For publishers and licensors, a ruling for Google could blunt unauthorized reuse of curated or licensed material; for toolmakers and smaller engineers, it could mean turning toward licensed feeds, partnerships, or a rethink of business models that have depended on scraping. For the web’s power dynamics, the case is a reminder that technical capability and legal right don’t always move in lockstep.