Court Dismisses Google's DMCA Suit Against Web Scraper SerpApi

Last week, a court dismissed Google's lawsuit against SerpApi, a web-scraping company. The judge ruled that Google has no grounds to sue under the DMCA: the corporation doesn't own the contents of its search results and hasn't shown that it's acting on behalf of rights holders. Throwing out a suit at such an early stage is a rare move, experts note.

A Strange Use of the DMCA

Google filed the suit in December, accusing SerpApi of circumventing anti-scraping measures and of selling search-results data through an unauthorized “Google Search API” service. It was an unusual gambit: search results in and of themselves aren't eligible for copyright protection. According to the company, that protection was safeguarding licensed content that appears in what are known as “knowledge panels.”

Google took its cue from a similar suit filed by Reddit in October against SerpApi and Google rival Perplexity. Reddit complained that the scraper was bypassing two layers of protection at once: the platform's own restrictions and Google's protection against scraping Reddit content from its results.

Meredith Rose, an attorney at the nonprofit Public Knowledge, calls this use of the DMCA “quirky” — the law was never designed for it. But it's hardly surprising: historically, the DMCA has been a handy tool for quickly cracking down on unwanted uses of content and forcing licensing negotiations.

A Narrow Path for Google

The judge pointed out that only a copyright owner, an exclusive licensee, or the party deploying the protective technology can bring a DMCA claim. In Rose's view, Reddit doesn't fit any of those definitions, so the ruling in Google's case bodes ill for it — the hearing on its own suit took place the same week.

Google was given 21 days to amend its complaint. The company intends to prove that rights holders expressly authorized it to apply the anti-scraping technology. But there's a risk here too: if Google argues that the “knowledge panels” are full of protected content, it would effectively be admitting that it reproduces unlicensed material, exposing itself to new suits. Rose thinks the corporation has “painted itself into a corner.”

The Battle for the Open Internet

SerpApi is defiant, calling the attack a “misguided attempt” to fence off the open internet by retroactively claiming rights to content that neither Google nor Reddit created. The scraper was sharper about Reddit: the platform, it said, wants to become a “toll collector” and monetize its users' content rather than protect it.

SerpApi's clients include Nvidia, Uber, and Adobe, all of whom rely on structured access to search data. The company says its business continues to grow despite the litigation.

Rose points to a broader problem: since the “API apocalypse” of summer 2023, publishers have been shutting off access to the open web en masse, which hurts not only AI training but also scientific research, archiving, and journalism. She calls the outcome of this “re-enclosure” of the web alarming.

A curious detail: Reddit's largest shareholder is Advance Publications, owner of Condé Nast — which also owns Ars Technica itself.

Source: Ars Technica