Google's Legal Action Against SerpApi: What It Means for Your Integration

If you build on SERP data, you have probably seen the discussion: in December 2025 Google sued SerpApi over the scraping and resale of its search results. It generated a lot of threads, a lot of speculation, and a fair amount of unnecessary alarm. It has also since been ruled on, though it is not over — on 20 July 2026 the court dismissed Google's central theory with prejudice, allowed a narrower one to be repleaded, and Google filed that narrower complaint on 10 August 2026. The case remains live. This is a calm walkthrough of what is actually known, what it does and does not change for you, and the one engineering habit that makes the whole question much less stressful.

Where the case actually stands

Updated 13 August 2026. This section was written in May 2026, when the case was at the filing stage. It has since been ruled on and then amended, so the original wording is superseded.

Google sued SerpApi in the Northern District of California (Google LLC v. SerpApi, LLC, No. 4:25-cv-10826), claiming that scraping its search results circumvented SearchGuard, its anti-bot technology, in breach of the DMCA's anti-circumvention provisions (17 U.S.C. § 1201).

On 20 July 2026, Chief Judge Yvonne Gonzalez Rogers granted SerpApi's motion to dismiss — but split the claim in two rather than throwing out the case wholesale:

• Dismissed with prejudice (no leave to amend) to the extent the claim was premised on SearchGuard controlling access to search results that do not contain any copyrighted content. The reasoning is the part with broad significance: § 1201 protects technological measures that control access to a copyrighted work, and where the results are plain URLs, snippets and factual index data, there is no protected work for a measure to guard.
• Dismissed with leave to amend to the extent the claim was premised on results that do contain a copyrighted component. Here the defect was narrower and fixable: Google had not alleged that SearchGuard was implemented and functioning to control access to those components with the authority of the copyright owners, which § 1201 requires.

Google took that second door. On 10 August 2026 — the last day of its 21-day window — it filed an amended complaint. The amended pleading reasserts claims under § 1201(a)(1)(A) and § 1201(a)(2), but aims at a much narrower target: Knowledge Panels containing licensed imagery, and the licence terms Google says authorise it to protect that material. It adds no non-copyright claims. Discovery reportedly remains stayed pending resolution of any motion to dismiss the amended complaint.

So the case is live, not over. Two things are true at once, and coverage tends to report only whichever half suits its headline: Google definitively lost the sweeping theory — it cannot use the DMCA to treat ordinary public search results as copyrighted works, and that holding is not coming back in this case — while a much smaller claim about licensed content inside Knowledge Panels is still being litigated.

We are intentionally not characterizing either party's legal position beyond what the docket shows, and none of this is legal advice. Read the primary sources rather than secondhand summaries, including this one.

Why this is not a reason to panic

Three reasons to keep a level head:

• A filing is not a ruling — and a ruling is not always the end. Nothing about your integration changed the day the suit was reported, the day it was dismissed, or the day it was amended. Service continuity and contractual terms are what affect you operationally, and those move on their own timelines.
• This space has been litigated before. Legal questions around scraping publicly accessible data are genuinely contested and have produced mixed outcomes over the years. One matter does not settle the field.
• Panic decisions are usually worse than the risk. Ripping out a working integration overnight introduces its own bugs and downtime. A measured evaluation beats a reflex.

The real lesson: provider risk is normal

The useful takeaway has little to do with this specific case. SERP data is infrastructure, and any single infrastructure provider — for any reason, legal or commercial or technical — can change pricing, availability, rate limits, or terms. That was true before this news and will be true after it resolves, whichever way it goes.

Teams that treat one SERP vendor as an unremovable dependency are carrying a risk that has nothing to do with lawsuits specifically. The mitigation is ordinary good engineering: make the provider swappable.

How to make any SERP integration swappable

Put a thin adapter between your application and whichever provider you use, so switching is a configuration change rather than a refactor:

# A single function your app calls. Swap the body to swap providers.
import requests

def search(query, country="us"):
    """Returns a normalized list of {position, title, url, snippet}."""
    resp = requests.get(
        "https://apiserpent.com/api/search",
        params={"q": query, "country": country, "apiKey": API_KEY},
        timeout=20,
    )
    resp.raise_for_status()
    return [
        {
            "position": r["position"],
            "title": r["title"],
            "url": r["url"],
            "snippet": r.get("snippet", ""),
        }
        for r in resp.json()["results"]["organic"]
    ]

The point is not which provider is in the function body — it is that the rest of your codebase never knows or cares. Pair this with: basic monitoring/alerting on the search path, a documented fallback provider, and one or two alternatives you have actually tested (not just bookmarked).

The provider landscape, briefly and neutrally

If you do decide to evaluate options — on a normal timeline, not a panicked one — the realistic field includes SerpApi (mature, broad engine coverage), Serper and SearchApi (simple, fast JSON), DataForSEO (cheap at scale, async), and Serpent API (multi-engine across Google, Bing, Yahoo, DuckDuckGo and Brave, flat per-call Quick Search pricing). Each has real tradeoffs; the right answer depends on your volume, latency needs, and feature requirements, not on a headline. We cover the comparison in more depth in our SerpApi alternatives guide and Google Search API alternatives guide.

Bottom line

This article is general information, not legal advice. For your specific situation, consult a qualified attorney.

FAQ

Was Google's lawsuit against SerpApi dismissed?

Partly — and the part that was dismissed is final. On 20 July 2026 Chief Judge Yvonne Gonzalez Rogers granted SerpApi's motion to dismiss. Where the results behind Google's SearchGuard measure contain no copyrighted content, the DMCA claim was dismissed with prejudice: plain URLs, snippets and factual index data are not a copyrighted work, so there is nothing for a technological measure to guard. Where the results do contain a copyrighted component, the dismissal came with leave to amend — the defect there was narrower, in that Google had not alleged SearchGuard was deployed with the authority of the copyright owners — and Google amended on 10 August 2026, the last day of its window. So the sweeping theory is dead in this case, while a much narrower claim about licensed imagery in Knowledge Panels is live and being litigated, with discovery reported to remain stayed. Nothing here is a final outcome.

Does the ruling mean scraping Google is now legal?

No, and reading it that way is the most common mistake in the coverage. The court decided one statute — the DMCA's anti-circumvention rules at 17 U.S.C. § 1201 — in one district, on one set of facts. Contract law, terms of service, computer-misuse statutes, trespass claims and privacy law are all untouched by it, and a decision in the Northern District of California does not bind courts elsewhere. The fair reading is “the DMCA is not the right tool for this,” not “anything goes.” This is general information and not legal advice.

Should I immediately stop using SerpApi because of the legal news?

Not as a panic move. The case has narrowed considerably in SerpApi's favour: the theory that scraping ordinary, non-copyrighted search results violates the DMCA is gone with prejudice, and what remains is confined to licensed imagery in Knowledge Panels. Nothing here obliges you to change vendors. The sensible response is the resilience practice that was always good engineering — keep your SERP provider behind a thin abstraction so you can switch if you ever need to. Whether you switch should be a normal evaluation decision, not a reaction to a headline.

Is scraping Google search results legal?

It is genuinely contested and varies by jurisdiction, but one narrow question got a clear answer in 2026. In Google LLC v. SerpApi, the Northern District of California dismissed with prejudice Google's claim that circumventing anti-bot measures to reach ordinary search results breaches the DMCA, holding that plain URLs, snippets and factual index data are not copyrighted works a technological measure can protect. That addresses one statute in one district. It does not make scraping lawful in general: contract, terms of service, trespass and privacy law are all still in play. If your business depends on search data, ask a qualified lawyer about your specific situation rather than relying on a blog post — including this one.

What is the practical takeaway for developers?

Treat SERP data as infrastructure. Any single data provider can change pricing, availability, or terms for reasons that have nothing to do with you. Design your integration so the provider is swappable, monitor the search path for disruption, and keep one or two evaluated alternatives ready. That advice held before the ruling and it holds after it, whichever way the remaining claim goes.