Google takes scraper to court, escalating the fight over who owns the web

Google has gone on the offensive against a web scraping company, it says has been quietly siphoning off its search results at an industrial scale. In a newly filed lawsuit, Google accuses SerpApi of illegally bypassing its protections to copy and resell search data, including copyrighted material from Google’s partners.

At the heart of the dispute is control. SerpApi markets itself as a service that delivers Google search results to developers, often for use in analytics and AI tools. Google says that access comes at a cost: the company claims SerpApi used deceptive techniques to evade bot-blocking systems and make automated requests look like they were coming from real people.

According to the complaint, Google introduced a technical safeguard called SearchGuard earlier this year to stop large-scale scraping of its search pages. That worked briefly. Google alleges SerpApi then engineered ways around it, disguising hundreds of millions of automated queries so they appeared to be normal human searches. In legal terms, Google says this amounts to unlawful circumvention of a technological protection measure under US copyright law.

Why does Google care so much? The company argues that its search results are not just a list of links. They often contain images, previews and other material licensed from third parties, shown in features like knowledge panels. Google says it pays for those rights, while SerpApi effectively republishes the same content without bearing any of those costs, undermining its agreements with content owners.

This is not SerpApi’s first legal headache. In October, Reddit sued SerpApi and other scraping firms, alleging that its content was being harvested and reused without permission, including for AI-related purposes. Google’s lawsuit does not directly name AI companies, but it lands squarely in the same debate about how data flows from the open web into commercial AI systems.

Google is asking the court to force SerpApi to stop bypassing its safeguards and to destroy any tools used to do so. That is a strong remedy, and it signals that Google wants to draw a clear line around what it considers acceptable access to its services.

Zooming out, this case fits into a much bigger reckoning. As AI models hunger for fresh, high-quality data, the incentives to scrape the web have exploded. At the same time, platforms like Google, Reddit and publishers are pushing back, arguing that unrestricted scraping threatens their business models and legal obligations.