Fraudulent DMCA Takedowns

What a fraudulent DMCA takedown actually does

A fraudulent DMCA takedown is a forged or knowingly false copyright complaint, filed under the Digital Millennium Copyright Act, whose purpose is to have a competitor's page removed from search results rather than to protect anything the filer owns. A takedown is the removal an online intermediary performs on receipt of such a complaint. Under 17 U.S.C. 512 that intermediary keeps its own liability shield by acting on a facially complete notice first and adjudicating the merits afterward, if at all.

That sequencing is the entire attack, and it is why this page carries the verdict it does. Every other attack vector documented on this site - the route by which one party degrades another's search visibility - has to persuade a ranking system of something, and ranking systems have spent fifteen years getting harder to persuade. This one persuades nobody. It skips ranking altogether. The reported URL is dropped from search results outright: it no longer appears for any query, including a search for its own exact title.

Nothing happens to the page itself. It still exists, still resolves, still returns a normal response to anyone who already has the address. What it loses is presence in the index - the property of being findable at all - and presence is not a thing you can argue an algorithm back into. There is no link to disavow, no spam signal to clean up, no quality problem to fix.

Two variants are documented, and they demand different evidence from you. In the first, the filer simply asserts that your page copies material they own, pointing at an unrelated or freshly created page as the supposed original; no copying happened at any point. In the second - the harder one - the filer copies your page onto a property they control, arranges for that copy to carry an earlier apparent date, and then reports your original as the infringement. A reviewer glancing at two near-identical pages and two dates cannot dismiss that on the face of the notice.

Note also which door the notice goes through. A complaint filed with a search engine removes the listing. A complaint filed with the host - the company whose servers actually serve your files - removes the page. Both routes exist, both are abused, and the second is worse because it takes the site down rather than merely hiding it.

Why this one does not have to fool an algorithm

Google's own copyright-removal documentation concedes, in its own words, the two facts that make the attack viable. It says Google removes search results linking to infringing content "when it is brought to our attention, and we do it quickly," and reports an average processing time of roughly six hours. It then states that Google "cannot always notify a site owner before content is removed" and "is not always able to verify the accuracy of a request" (Google, Search removals due to copyright infringement). Review is not absent - Google says it declines requests for URLs that clearly did not infringe - but it is not reliable enough to catch a competently drafted forgery at that speed and that volume.

The 2026 record is unusually probative, because the victims were the last people who should have been vulnerable. In March 2026 Search Engine Land had an investigative article removed from Google Search globally after a complaint alleging the piece "copied our entire content word for word" and used "proprietary images." The article contained no images at all. It was reinstated on 31 March 2026, days later (Danny Goodwin, Search Engine Land, 30 March 2026). In June 2026 the same thing happened to Press Gazette, whose article about AI reporters generating error-strewn stories for football websites was delisted on a notice citing, as the supposed original, a month-old forum post about online casinos. It came back on 2 July 2026.

Two trade publications with newsroom lawyers and direct lines into Google lost pages for days. A business without a masthead should plan for longer.

The strongest single piece of evidence is not a victim's account at all - it is Google's. In November 2023 Google filed suit in the Northern District of California against two individuals it alleged had created at least 65 accounts to file fraudulent notices against more than 117,000 third-party URLs, in order to clear competing sellers out of search results in a single retail niche. Google's complaint states that the defendants "have weaponized copyright law's notice-and-takedown process and used it not for its intended purpose of expeditiously removing infringing content, but instead to have the legitimate content of their competitors removed based on false allegations." Investigators traced notices touching around 500,000 URLs in total. I have not verified the disposition of that case and do not state one here; the filing itself is the point. The intermediary that operates the process went to federal court to describe it being used as a competitive weapon at industrial scale.

Nothing has changed in the victim's favor. There is no equivalent here of the shift from demotion to devaluation that defused most link attacks, because a copyright removal is a compliance action rather than a quality judgment - there is no signal to discount. If anything the picture has worsened: notice volume has grown past the point where per-notice human scrutiny is economic.

The signature: indexed, but unfindable

This vector is misdiagnosed more often than any other on this site, and always in the same direction, because the report everyone checks first is silent.

• Manual actions is empty. A copyright delisting is not a manual action - not a penalty applied by a human reviewer at Google and shown to you in Search Console - so it never appears there. An owner who checks that one screen, sees it clean, and concludes the drop must be algorithmic has just talked themselves out of the correct diagnosis.
• Indexing reports look normal. The URL usually still reports as indexed, because the suppression happens when results are served, not when pages are crawled or indexed. Indexed but unfindable is the signature. No other cause produces that pair.
• The footer of the results page. Search a distinctive verbatim sentence from the missing page. "In response to a complaint that we received under the US Digital Millennium Copyright Act," Google's notice says, "we have removed results from this page." It sits at the very bottom of the results, usually with a link to a public copy of the notice. Almost nobody scrolls that far, which is why this goes undetected for weeks.
• The Lumen Database. Google states that it shares a copy of qualifying copyright removal requests with Lumen, where site owners can inspect them. That copy is where you obtain the notice text, the URL claimed as the original, and the submitter's stated identity - the evidentiary core of anything you do next. Search it by your own domain.
• The Transparency Report. Google's Transparency Report - its standing public accounting of removal requests, government demands and traffic disruptions - lets you browse copyright complaints by the domain they name. It confirms scope: one URL, or a systematic run at a section.
• Analytics. Near-total loss of organic traffic to one URL or a small handful, with the rest of the site untouched. Algorithmic movement does not look like that.

It gets mistaken for a core update, a manual action, an accidental noindex, a robots.txt error, a canonical misconfiguration and a hosting outage. The test that separates it from all six takes two minutes: the footer notice, plus a clean manual actions report.

The counter notification, and the clock it starts

Absent a counter notification the delisting is permanent. This is one of the few pages on this site where doing nothing is the wrong answer, and I want to be blunt about it: waiting does not restore anything, and the longer you wait the more of the traffic is gone for good.

Section 512(g)(3) sets out what a counter notification must contain: your signature; identification of the removed material and the location where it appeared before removal; a statement under penalty of perjury that you have a good faith belief the material was removed as a result of mistake or misidentification; and your name, address and telephone number, together with a statement consenting to the jurisdiction of the federal district court for your district - or, if you are outside the United States, any district in which the provider may be found - and agreeing to accept service of process from the complainant.

Then the clock. On receipt of a valid counter notification the provider forwards it to the original complainant and must restore the material "not less than 10, nor more than 14, business days following receipt of the counter notice", unless the complainant first tells the provider it has filed a court action to restrain you. Read that as it is written: ten business days is a floor, not a target. Roughly two calendar weeks is the fastest the statute permits, and nothing you do accelerates it.

Three consequences that people are rarely told before they file:

1. Your name, postal address and telephone number are forwarded to the person who attacked you. That is a statutory requirement, not a platform choice, and it is the single most common reason a legitimate victim declines to file.
2. You are consenting to United States federal court jurisdiction. For a business outside the US that is a real decision with real exposure, not a formality on a form.
3. A false counter notification carries the same liability as a false notice. The misrepresentation provision runs in both directions. If you are not certain the material is yours, get advice before you swear to it.

Before you file, preserve the proof of authorship you will need if the complainant does sue inside the window: original publication timestamps, CMS revision history, server logs, archive captures of your page predating the claimed original, and the author's drafts. Priority of publication is the whole case. Collect it first, because remediation destroys evidence and you only get one pass at it.

Escalating publicly is worth attempting in parallel, but it is not a plan. Both 2026 press cases were reinstated after their own newsrooms wrote about them. That route was available because the victims were journalists.

What the perjury declaration actually covers

This is the detail that explains why the process is so easy to abuse, and it is almost universally misread. A complying notice must include, under 512(c)(3)(A)(v), "[a] statement that the complaining party has a good faith belief that use of the material in the manner complained of is not authorized by the copyright owner, its agent, or the law." Separately, under (c)(3)(A)(vi), it must include "[a] statement that the information in the notification is accurate, and under penalty of perjury, that the complaining party is authorized to act on behalf of the owner of an exclusive right that is allegedly infringed."

Look at where the oath attaches. The penalty of perjury covers the filer's authority to act for the owner. It does not cover the accuracy of the infringement claim. A person who genuinely owns the imaginary original they are asserting has sworn to nothing false in the eyes of that clause, and that drafting is a large part of why perjury prosecutions over false notices are effectively unheard of.

So when somebody reassures you that the filer swore under penalty of perjury and therefore would not have risked it, they have read the wrong sub-clause. The deterrent in the statute is not the oath. It is 512(f), and 512(f) is weaker than it looks.

Section 512(f), and what it is realistically worth

Section 512(f) makes any person liable who "knowingly materially misrepresents" that material is infringing, for damages, costs and attorneys' fees incurred as a result of the provider relying on the misrepresentation. Two decisions define what that is worth in practice, and they point in opposite directions.

Online Policy Group v. Diebold, Inc., 337 F. Supp. 2d 1195 (N.D. Cal. 2004), decided 30 September 2004, was the first successful application of the provision. The court construed the terms broadly: "knowingly" means "a party actually knew, should have known if it acted with reasonable care or diligence, or would have had no substantial doubt had it been acting in good faith, that it was making misrepresentations," and "material" means the misrepresentation "affected the ISP's response to a DMCA letter." The court found the defendant had knowingly materially misrepresented infringement, observing that no reasonable copyright holder could have believed the material at issue was protected. The case settled on 15 October 2004 for $125,000 in damages and legal fees.

Lenz v. Universal Music Corp., 801 F.3d 1126 (9th Cir. 2015), decided 14 September 2015, cut the other way for anyone hoping to use the provision. The Ninth Circuit held that copyright holders must consider fair use before sending a notice - genuinely helpful - but also held that the good faith belief requirement "encompasses a subjective, rather than objective standard," requiring "some actual knowledge of misrepresentation on the part of the copyright owner." A filer who can plausibly claim sincere, if unreasonable, belief escapes. The Supreme Court declined to review it on 19 June 2017.

Commentators disagree about the net effect and I am not going to pretend the disagreement is settled: one camp reads Diebold as establishing real exposure, the other reads Lenz as having gutted the provision. Both readings are defensible, and which one governs your situation depends on a fact you can assess yourself.

Here is the assessment worth having. The commentary that treats 512(f) as a dead letter was written about overzealous rights-holders - people who genuinely own something and overreach. Your attacker is not that. Where the claimed original is fabricated or backdated, actual knowledge of misrepresentation is inferable from the fabrication itself, because nobody backdates a page by accident. The subjective standard is a much lower obstacle in the fraud scenario than in the scenario the case law was built on.

The real obstacles are not legal, they are practical: identifying who filed, and reaching them. Notices arrive from disposable accounts, frequently from outside US jurisdiction. Where the filer is identifiable, counsel may also want to look at tortious interference with business relations and at state unfair-competition statutes.

Who is genuinely most exposed

The mechanism works against any indexed page, but the damage is wildly unequal, and knowing where you sit tells you how hard to fight.

• Single high-value pages. A business whose revenue depends on two or three commercial URLs can lose most of its organic pipeline to a notice against one of them.
• Time-sensitive content. News, investigation, launch coverage, anything seasonal. Restoration after two weeks returns the URL and not the traffic. That traffic was spent while the page was invisible, and it does not come back later.
• Owners who do not read Search Console messages, or who have no alerting on per-URL organic traffic. The detection lag is usually longer than the statutory remedy.
• Businesses outside the United States, for whom the counter notification requires consenting to a foreign court's jurisdiction. Attackers know that hesitation exists and some are counting on it.

And the group that should worry least: a site with hundreds of substantial pages and no single dominant money page. A notice against one URL there is an irritation with a fixed, known remedy, not an emergency. If someone is selling you continuous monitoring against copyright complaints, ask what it would do that a weekly check of per-URL organic traffic and a quarterly look at the public notice databases would not. Usually the answer is nothing.

What not to do while the page is gone

Every item on this list is something I have seen done, at cost, by someone who had correctly noticed a problem and then reached for the nearest familiar tool.

• Do not disavow anything. Links have no role in this mechanism whatsoever. A disavow file cannot restore a delisted URL and carries its own risk if built carelessly.
• Do not file a reconsideration request. Reconsideration is the appeal against a manual action, and there is no manual action here. The request goes into a queue and is answered, correctly, with the observation that nothing has been applied to your site.
• Do not rewrite the page to look less like the claimed original. This is the most damaging response available. It reads as an admission, and it destroys the record of what you actually published - which is the evidence your counter notification rests on.
• Do not delete the page, noindex it, or block crawlers. None of these touch a removal that operates on a URL at serving time, and each one costs you something real.
• Do not republish the same content at a new URL and hope to slip past. You now have two URLs with the same problem and a weaker claim to either.
• Do not retaliate with a notice of your own. Filing a complaint you know to be false is the same knowing material misrepresentation, with you on the wrong end of it.

The correct sequence is short: confirm the cause from the results-page footer and the public notice copy; preserve the notice and your own proof of authorship; file the counter notification; wait out the statutory window; and only then decide whether the filer is worth pursuing.

Frequently asked questions