🔓 WordPress Publishers Just Got a New AI Scraping Reality Check

Google's attempt to treat an anti-scraping system as copyright protection failed in court, exposing a legal assumption many publishers never questioned.

Chief Judge Yvonne Gonzalez Rogers dismissed Google's DMCA claims against SerpApi on July 20.

For WordPress publishers relying on anti-bot protections, the important part wasn't who won. It was why.

SerpApi's entire business is scraping Google's search results and reselling them through an API, increasingly to AI companies. Google argued SerpApi broke the law by defeating SearchGuard, the internal system Google uses to detect and block automated scraping.

The judge disagreed. Her reasoning reaches far beyond Google.

SearchGuard protects Google's ad revenue, not a copyrighted work, the judge found. DMCA anti-circumvention law exists to protect copyright, not business models.

A wall built around your revenue stream is not the same thing as a lock on a copyrighted file, and the court threw out Google's claim with prejudice everywhere no copyrighted content was involved.

Google gets 21 days to come back with a narrower claim limited to Knowledge Panel images.

Neither party here deserves sympathy.

SerpApi scrapes at industrial scale and resells the output, feeding the same AI companies publishers are trying to fend off.

Google built a trillion-dollar business by crawling everyone else's pages for two decades, then wanted copyright law to stop someone from crawling its own.

This is not a story about good guys. It is a story about what happens when a legal assumption publishers have relied on gets tested in court, and the answer just changed.

Most WordPress publishers run some equivalent of SearchGuard: an anti-bot layer, paywall protection, rate limiting, or CDN bot management. These tools are designed to protect revenue, not a specific copyrighted work.

Before July 20, plenty of site owners assumed the DMCA gave them a legal backstop if someone defeated that layer.

This ruling says otherwise: circumventing a wall that protects your business model, rather than a copyrighted work specifically, does not trigger DMCA anti-circumvention liability.

That is the risk transfer.

The legal tool many publishers assumed was sitting in reserve turns out not to cover the exact scenario they built it for. A parallel case, Amazon v. Perplexity, is testing the same question from a different statute, asking whether an AI agent counts as an authorized visitor under the CFAA when it acts on someone's website.

Different law, same underlying fight: on what terms is an automated visitor allowed onto content you publish?

Courts are drawing that line case by case, and none of the parties currently in front of a judge are individual site owners.

The legal terrain here is still moving, and it might not move in your favor. What does not depend on a judge is the technical layer you already operate.

Robots.txt, host-level bot management, and CDN-level crawler rules remain yours to configure regardless of how DMCA circumvention law shakes out.

WordPress site owners can decide, crawler by crawler, what gets in and what gets blocked, using the controls their host or CDN already provides.

That decision does not require winning a lawsuit. It requires making the call before someone else makes it for you, whether that is a court refereeing a fight you are not part of, or a plugin default you never reviewed.

The open web runs on a consistency problem. You cannot demand open access to everyone else's content for your own discovery and citations, then claim a legal shield the moment a machine you dislike reads yours.

This ruling does not ask you to like that trade. It asks you to stop assuming the law will make the choice for you.

Which crawlers are you blocking right now, and did you choose that list on purpose, or did you inherit it from a default someone else set?

Read the story →

This week in WP Radar: AI visibility myths agencies keep repeating, a class action surviving against Automattic, AI threatening publishers' business model, and AI slop competing for ad dollars.

Google's AI Overviews now appear in 43% of searches, up from 15% a year ago.

Similarweb data shows AI Mode visits more than doubled to 279 million by May 2026, yet only 6.8% of U.S. ChatGPT desktop queries carried a citation as of that month.

A May update did push cited-page referral share from 25% to nearly 60%, but the underlying shift, Google becoming the destination instead of the doorway, is not reversing. Publishers who track only total traffic are measuring a shrinking share of the story.

Read the report →