The Ruling

On 20 July, a federal judge dismissed Google's DMCA lawsuit against SerpApi, a company that scrapes Google search results and resells them as structured data. Judge Yvonne Gonzalez Rogers ruled that plain search results—URLs, snippets, and index data—are public facts, not copyrighted works. Therefore, Google's anti-bot measures (SearchGuard) are not protected under DMCA Section 1201, which only covers technology guarding copyrighted content.

"Google has not alleged a plausible violation of the DMCA," the judge wrote. The claims were dismissed with prejudice (cannot be refiled).

Google's Narrow Path

Google has 21 days to refile on a smaller claim: that licensed snippets in knowledge panels are copyrighted. But that argument could backfire. Meredith Rose of Public Knowledge notes that Google does not license everything in knowledge panels. If Google claims those panels are copyrighted, it invites lawsuits from rights holders whose content appears in Google's AI Overviews.

"They have talked themselves into a little bit of a corner," Rose told Ars Technica.

What's at Stake

Search data is critical for AI training. Google offers no official search API, so firms like SerpApi are the main route to Google's index. Customers include Nvidia, Uber, Adobe, and Perplexity. SerpApi's CEO Julien Khaleghy noted that Google's own damages calculation, if applied literally, would exceed the entire US economy.

Reddit's Parallel Suit

Reddit filed a similar DMCA suit against SerpApi and Perplexity in October. Its dismissal hearing followed Google's loss. The same legal gap applies: Reddit is neither copyright owner nor exclusive licensee of user content in search results.

Implications for Developers

  • Scraping public data is legal if the target has no copyright on the data. This ruling reinforces that factual data (URLs, snippets) is not copyrightable.
  • DMCA cannot be used to block scraping of non-copyrighted data. Anti-bot measures are only protected if they guard copyrighted works.
  • AI training data access may remain open. Companies scraping for AI training can cite this precedent.
  • Expect more litigation as publishers and platforms try to wall off data. The web's open nature is being contested case by case.

Next Steps

Google plans to file an amended complaint. Developers should watch for the refiled claim on knowledge panels. If Google wins that, it may strengthen platform control over structured data. If it loses, scraping public web data becomes safer.

Technical Context

SerpApi uses browser fingerprint spoofing, IP rotation, and CAPTCHA solving to bypass SearchGuard. The court agreed these are circumvention, but not illegal without a copyrighted work behind the barrier.

For developers building scrapers: focus on extracting factual data (URLs, public snippets) rather than copyrighted content like images or long-form text. Use rotating proxies and respect robots.txt to avoid legal risk, though this ruling reduces DMCA liability for factual data.

The Bigger Picture

The "re-enclosure" of the web began around 2023 with aggressive AI scraping. Publishers are walling off content, but that also breaks anonymous crawling for research, archiving, and journalism. This case sets a precedent that public facts remain free to access at scale.