Google Loses DMCA Suit Against SerpApi Over Scraped Data

A Setback for Google in the Data-Scraping Wars

A federal judge has dismissed Google's DMCA lawsuit against SerpApi, a company that scrapes and resells Google search results data, ruling that copyright law cannot be used to prevent the collection of factual, uncopyrighted information 1. The decision underscores a long-standing legal principle: raw facts, including the contents of a search results page, are not protected intellectual property, even if the way they're compiled or presented might be 1. For Google, which has increasingly leaned on legal tools to police how its data is harvested and reused, the ruling is a reminder that copyright law has real limits when it comes to information that is inherently public.

Part of a Much Bigger Legal Reckoning

The SerpApi dismissal lands amid an escalating wave of litigation over how data — copyrighted or not — is scraped, licensed, and monetized in the age of generative AI. Three major publishers have separately moved against Google itself, accusing the company of using their journalism to train AI systems without compensation, the latest salvo in what has become a sustained campaign by media companies to extract payment from AI developers 2. News Corp has taken a similar fight to Brave, suing the browser maker over alleged AI scraping practices, a case that illustrates how publishers are now weighing whether to sue AI companies outright or instead negotiate licensing deals as a more pragmatic path forward 3.

The battles extend well beyond text and news. Sony has filed a second copyright lawsuit against AI music startup Udio, this time alleging that more than 30,000 sound recordings were used without authorization to train Udio's music-generation models, a sign that entertainment companies are not backing down even after initial legal skirmishes 4.

The First Real Money Changes Hands

While most of these disputes remain unresolved, one has already produced a landmark outcome: Anthropic's $1.5 billion settlement over AI training data has now received final court approval, marking what appears to be the first substantial payout to emerge from the broader wave of AI copyright litigation 5. That settlement offers a concrete benchmark for plaintiffs and defendants alike, suggesting that even industry-defining AI companies may find it more prudent to negotiate large settlements than to risk unpredictable jury verdicts or prolonged appeals.

Why It Matters

Taken together, these cases sketch a legal landscape still very much in flux. Courts are drawing distinctions between scraping uncopyrightable facts, as in the SerpApi case, and using copyrighted creative works — news articles, music recordings — to train commercial AI models. Publishers, record labels, and tech companies are testing multiple strategies simultaneously: lawsuits, licensing negotiations, and settlements. With Anthropic's payout now finalized and fresh suits piling up against Google, Brave, and Udio, the outcomes of these parallel fights will likely shape how AI companies source training data — and how much they'll have to pay for it — for years to come.