Someone Filed a Bogus DMCA Notice to Kill a Story About A Sketchy SEO Firm. It Worked — Briefly.
from the gee-who-could-have-the-incentive-to-do-that? dept
We’ve talked for years about how the DMCA’s notice-and-takedown system is ripe for abuse. The legal structure of the law practically begs for such abuse: send a notice, content disappears, and the target has to fight through a slow counter-notice process to maybe get it back. The system rewards speed of takedowns over accuracy because the burden of getting it wrong really only works one way. Sites have incentive to take content down first and ask questions later to avoid facing expensive liability. Getting it wrong may frustrate those whose content has disappeared, but there’s basically no legal cost to the platform. But if they get something wrong and leave infringing content up, they could face a very expensive legal bill. Which means anyone with something to hide and no particular attachment to honesty has a ready-made censorship tool at their disposal.
And while Google is rare in that it is much more aggressive in rejecting DMCA notices than most other sites, that doesn’t mean that it’s perfect.
Last week, Press Gazette published an investigation into Clickout Media, a UK-based company that has been buying up respected online news outlets, gutting their newsrooms, replacing human journalists with AI-generated writers (complete with AI-generated profile photos), and stuffing the sites full of affiliate links to offshore gambling operations. The whole game is to exploit the acquired sites’ reputations and search rankings — what’s known as “parasite SEO” — to drive gambling traffic through what look like legitimate publications. It’s a really excellent piece of reporting about a practice that is gutting digital news brands. Just a quick snippet, though it’s worth reading the whole thing:
Speaking anonymously, one former Clickout Media employee said: “I was moved from site to site. Writing guidelines and strategies changed every other week with very little explanation. At first, I didn’t write casino content, but then I wrote articles on bets and odds. Then AI articles started appearing.”
The owners of one site bought by UK-based Clickout Media said they were approached by anonymous buyers in the first instance.
The organisation has previously bought multiple sites in football and women’s sports (Football Blog, She Kicks, Sportslens, Sportslens UK, Sportscasting UK, Football Blog UK), as well as gambling sites, including Gambling Insider, for which it is suggested Clickout Media paid at least £12m.
However, within days of being published, the exposé disappeared from Google’s search results, removed after a DMCA copyright complaint.
A search of the exact Press Gazette headline: “The SEO parasites buying, exploiting and ultimately killing online newsbrands” does not bring the article up.
A note at the bottom of the Google search results page reveals for this query states: “In response to multiple complaints that we received under the US Digital Millennium Copyright Act, we have removed 2 results from this page. If you wish, you may read the DMCA complaints that caused the removals at LumenDatabase.org: Complaint, Complaint.”
A follow-up article by Search Engine Land covering the same topic was also removed (that’s the second “complaint” link above). So whoever was behind this was being thorough.
Now, Press Gazette doesn’t definitively identify who filed the takedown notice, and we should be careful here too. The complaint was filed anonymously from “US Hub” which gives us little info but which Press Gazette notes “suggests the complaint originated outside the US.” You can connect the dots yourself on who has a reason to make an investigative exposé of Clickout Media vanish from search results, but we can’t say for certain.
What we can say for certain is that the takedown notice itself is laughably, almost impressively, bogus. You can read it yourself over at the Lumen Database. The complaint claims Press Gazette’s entirely original investigation infringes on an unrelated article published by The Verge. That Verge article? It’s actually about Google cracking down on sketchy SEO practices, the likes of which Clickout Media seems to engage in. Which is, if nothing else, a spectacularly on-the-nose URL to attach to a fraudulent takedown of an article about sketchy SEO practices.
The language of the notice is quite something:
The infringing news website has blatantly and willfully violated copyright law by copying our entire content word for word, including all images, which are solely owned by our company. This includes the complete replication of our original written material, as published on our official website, along with the proprietary visuals accompanying it. Despite multiple good-faith efforts to resolve this matter amicably, the infringing party (hereinafter referred to as “Infringer”) continues to unlawfully publish and distribute our copyrighted content without permission. This is a direct and flagrant breach of our rights and a clear violation of Google’s copyright policies. We hereby demand the immediate removal of this infringing material from Google search results to protect our intellectual property.
None of that is true. Not one word. The Press Gazette article is original reporting. It has nothing to do with the Verge piece cited as the “original” work. There were no “multiple good-faith efforts to resolve this matter amicably,” because there was no infringement to resolve. The whole thing reads like someone fed a prompt into a chatbot asking it to write an angry-sounding but legally meaningless DMCA notice, and then pointed it at an article they wanted to disappear.
As the Press Gazette report on the bogus takedown notes, SEO experts found the whole thing bizarre, in part because Google is actually much better than most at sniffing out bullshit DMCA takedowns. But this one they missed.
Writing on X, SEO consultant Glenn Gabe said: “Surprised this was approved by Google…This is a BS DMCA takedown that doesn’t even make sense.”
Google processes an absolutely massive volume of takedown requests and rejects a good chunk of them. But this is the Impossibility Theorem in action: at that kind of scale, even a system that works well most of the time will let nonsense through sometimes. One bad notice that should have been caught in a ten-second review slips past, and suddenly a major piece of investigative journalism is invisible to anyone searching for it.
The good news is that, as of March 31, the Press Gazette article was reinstated in Google’s search results. The system worked, eventually. But that ‘eventually’ is doing a lot of heavy lifting. The article was invisible during what was probably the peak window of public interest in the story. Legal challenges to DMCA takedowns can take weeks or months to resolve, and the people who file these bogus notices know that. The copyright-enabled censorship just has to last long enough to blunt the impact.
For what it’s worth, the Press Gazette isn’t the only outlet digging into Clickout Media’s practices. Aftermath recently published its own extensive investigation based on eight months of reporting and interviews with more than two dozen current and former employees. That piece documents AI-generated author profiles, fake credentials (one supposed writer claimed an MA from Oxford in a program the university confirmed has never existed), writers being told never to publicly acknowledge any connection to Clickout Media, and a systematic strategy of acquiring beloved gaming publications only to fill them with crypto casino links.
Clickout Media is getting more and more negative attention, and Streisanding the Press Gazette story by having it removed from search will probably just attract more investigative reporters to the subject.
The company already has a pretty sketchy pattern: buy a respected publication, exploit its reputation, squeeze out whatever search ranking value you can, and discard the husk. And when someone publishes an article documenting what you’re doing, apparently get someone to file a bullshit copyright claim to make that article disappear too. It’s sketchy SEO all the way down.
This is why those of us who spend our time in the weeds of internet law won’t shut up about how legal liability systems are structured. The DMCA’s notice-and-takedown framework already gives bad actors a weapon to suppress speech. You don’t need a legitimate copyright claim. You don’t even need a coherent one. You just need to file the paperwork and wait for an automated system to do its thing.
And every time someone proposes weakening Section 230, or creating new obligations for platforms to proactively police third-party content, or imposing liability for hosting material that someone claims is harmful — they are, whether they realize it or not, proposing to hand bad actors this same kind of weapon in a dozen new calibers. The DMCA is the version of this we already have, and we can see plainly how it gets abused. We should be fixing the current system, and punishing the widespread abuses, rather than spreading that same broken incentive structure to every other area of online speech.
Bad actors will always exploit whatever legal lever is available to suppress content they don’t like. The question for policymakers is whether you’re going to keep handing them more levers. These kinds of bogus DMCA takedowns should be a warning for all those demanding reforms “weakening” Section 230. Because if you think bogus DMCA takedowns are bad now, just wait until they’re not just about copyright.
Filed Under: copyright, dmca, dmca takedown, parasite, search, seo, spam
Companies: clickout media, google, press gazette
Comments on “Someone Filed a Bogus DMCA Notice to Kill a Story About A Sketchy SEO Firm. It Worked — Briefly.”
But, as bad as the law is, it doesn’t actually allow for takedowns of references to copyright infringement (unless perhaps judges invented such a requirement via case law). Google apparently chose to allow people to use notices that way, despite a lack of any legal basis. And other search engines kind of copied from them. Maybe the film companies applied pressure as advertisers.
By contrast, in the old days, I don’t think anyone ever had their phone number removed from the phone book, or disconnected altogether, based purely on unproven accusations of illegal activity (like maybe a video store that was a little shady regarding copyright law). I’m not sure that was even an option after a court found someone guilty or liable.
My only objection is insisting that the DMCA’s problems are inherent to the notice-and-take-down framework, instead of acknowledging that there are a lot of different choices you make, that can meaningfully change the incentives. If there are meaningful penalties towards a bogus claim (like jail time), that’s a very different incentive structure. You can also do things like choose to keep the content up during litigation (DMCA already requires restoring it after 10-14 days of counternotice, absent that. Another knob to tune). And that is by no means a complete list.
The DMCA is not just an example of a notice-and-takedown framework. It’s an example of a very permissive one tilted towards the copyright holder, that doesn’t do much to stop bad actors. There’s an entire spectrum of tweaks you could make, and there’s no justification for asserting the DMCA is the only way it could possibly work.
But we can also see plainly how it could be fixed. Even in the DMCA alone, just fixing things like the borked perjury clause would go a long way. It’s there.. they just screwed it up and didn’t fix it. You can disincentivize bogus claims, if that were a priority.
As flawed as the DMCA is, I think it’s actually not too bad as a first crack at the problem. The real sin is not meaningfully updating obvious problems in over 2 decades.
Re:
Until those incentives are changed, the DMCA’s problems are inherent to the notice-and-takedown framework under which it operates. Good luck changing those incentives, though. Hollywood alone could (and would) destroy your effort without any effort.
Re: Re:
I think this is kind of mixing terminology. If something is inherent, it means it can’t be fixed/changed, it’s fundamental to how it functions regardless of a specific implementation. There are certain parts of notice-and-takedown that can’t be fixed, just because it’s notice-and-takedown.
Yeah, I mean, it’s not going to happen any time soon. But you have to start somewhere. If we want a better world, it’s gonna have to go through Hollywood. Realistically, you’re going to have to break the Hollywood lobby at some point or another, even if it’s for other things like copyright reform or whatever.
Re: Re: Re:
I mean, if you think you can change the DMCA, all power to you. But like I implied, you’re facing a massive uphill climb in that effort. In that sense—and in the sense that a notice-and-takedown system inherently favors bullshit notices because it requires a takedown as soon as a notice is received—the DMCA is a static system.
I’m on your side and all, but let’s be pragmatic here: If you think Hollywood—i.e., the movie/TV studios, the broadcast networks, the streaming services—is the only force you have to stand up to here, you’ve lost your crusade before you’ve even unsheathed your sword. A thousand castles each occupied by a thousand enemies who each have a thousand ways to squash you like a bug are what stand between you and your goal. I’m not saying you can’t fight; again, I am on your side. But consider that a threat of genocide from the sitting POTUS isn’t enough to make Congress rein in the war powers of the presidency. What makes you think Congress will ever stand up against every copyright lobby in the world?
Re: Re:
How can it be “inherent” when nothing in the DMCA allows for takedowns of links to infringing material?
Re: Re: Re:
Legally? No, it doesn’t allow for that. But tell that to people who lack the resources to fight back against bogus takedowns. Hell, I can’t think of many companies that would risk being held legally liable over content targeted by what would be, legally speaking, a valid takedown.
Re: Re: Re:
Because there is no mechanism or requirement for the filer of the notice to prove the notice is valid before issuing it. It is assumed that every notice is valid without evidence of its validity.
It’s the equivalent of if, in civil litigation, on filing a claim with the court the respondent is found liable immediately, and then the trial is held where the respondent has to prove they are not liable.
Re: Re: Re:2
Theoretically, the DMCA takedown process only removes liability. If ignored, the web host is in the same position as any web host would’ve been before 1998 (when the law was passed).
You’re right that the notices being treated as valid is a problem, but the law as written does not require that. So I don’t agree that it’s “inherent”; for all we know, some of the people who voted for that law could be horrified at how people are treating it.
Stephen’s also kind of right that no company seems to want to go “against the herd”. And, yet, before the DMCA created this safe harbor, pretty much every U.S. company hosting user-generated material managed to get along just fine.
Re: Re: Re:3
It kind of does, though? I mean, yes, a company could risk its protection from legal liability for third-party speech by refusing to comply with a takedown notice. But few companies would do that when lawsuits are a drain on resources even if those suits get dropped or punted out of court before getting any further than a motion to dismiss. The perverse incentive of the notice-and-takedown system lies in that calculus: Even if the speech goes back up after a takedown, the initial takedown is often the goal, and whether a notice is valid doesn’t matter because the courts have almost never punished people who file false notices.
And how long do you think that would’ve lasted without the DMCA, especially in the United States? As much as I don’t like saying it, we kind of needed the DMCA to codify those “safe harbor” protections so sites like YouTube wouldn’t get annihilated by the corporate copyright cartels. But it’s long past time to reform the notice-and-takedown system at the heart of the DMCA. Changing it to notice-and-notice would be a good start; so would adding more (and stronger) penalties for filing a false takedown notice.
Re: Re: Re:4
It’s a bad situation, but as stated, it used to be the normal situation. And it’s still the normal situation in other areas. I’m aware of no safe harbor for sexual images of children, or for “revenge porn”—especially internationally—and any company accepting image uploads is accepting the risk of such things. Courts seem to understand it would not be fair to let a company be sued into oblivion when they’re taking reasonable steps to comply with laws.
It’s hard to say, but I think this view is overly pessimistic of the courts. We didn’t have section 230 protection either, and courts decided site operators were not liable for libel and such as long as they didn’t make any editorial decisions. People consider that problematic, but things might have developed to the point where some moderation system not operated by the site operators would’ve been found acceptable, and maybe would’ve been extended to copyright.
The DMCA provided an easy but bad alternative, pretty much blocking all such development.
Re: Re: Re:5
And it was a precarious one, even with Section 230 intact. The DMCA made it less precarious, since a website could simply comply with the DMCA and not get a lawsuit dropped on it for daring to be, in a copyright holder’s view, legally liable for copyright infringement.
Those things are crimes. Copyright infringement is, far more often than not, a matter for civil courts.
I have good reasons to be pessmistic of American courts—and the courts have given me those reasons by often being nonsensical and ridiculous in regards to, among many other aspects of the legal system, copyright law. Deference to the copyright cartels is the norm, not the exception, until and unless the courts give me reason to think otherwise.
But prior to 230, courts had also ruled that yes, websites could be held liable for third-party speech if those sites moderated such speech. 230 changed the standard for liability. The DMCA made it possible for websites to allow third-party speech and not risk being held liable for copyright infringement under similar “they moderated [x], but not [y]” situations. It wasn’t the most elegant solution, but it did prevent that problem from growing out of hand and risking the existence of the open Internet.
That situation is, paradoxically, what 230 was designed to both enable and prevent: Under 230, a website can experiment with moderation that best serves its interests without it needing to take its moderation to one extreme (moderate all speech) or the other (moderate no speech) for the purposes of avoiding legal liability for that speech.
The DMCA takedown system is the system you’re talking about. Lots of websites have automated moderation tools; a copyright database check, like YouTube’s Contend ID, is one such tool. That kind of tool isn’t explicitly an automated DMCA system, but it’s close enough to one that it may as well be, especially when such checks can stop content from being viewed altogether even when the content is presented under Fair Use laws. James Stephanie Sterling invented the Copyright Deadlock—the use of multiple copyrighted works overlapping with one another—to prevent any one company from being able to shut down their work on YouTube and/or monetize it for anything more than what would amount to pocket change. Without the DMCA, the chances that Sterling would have been sued into oblivion for even Fair Use–qualifying uses of copyrighted material would be drastically higher—and the chances that YouTube would be held liable for those acts of infringement would also be much higher. Again, the DMCA isn’t perfect, but it was needed to keep the copyright cartels from wrecking the Internet. Now it needs reform, starting with the bullshit takedown system, and that’s to say nothing of copyright law in general needing even more reform to account for modern technology.
Re: Re: Re:6
I phrased that poorly. What I meant was that jurisprudence would probably have converged on something like common sense. Like, maybe site operators wouldn’t be liable for user behavior at all (provided they didn’t do their own “curation”), or maybe only if they respond to reasonable copyright complaints.
I don’t think it would have ended up like the DMCA takedown system, where complaints are presumed valid. If we look outside the USA, most other countries don’t have anything quite so strict, nor are they holding site operators responsible for everything.
Re: Re: Re:7
I’d love to have your optimism. Or is it naïveté? Either way, it’s a bit foolish to think the American legal system would operate with “common sense” in re: copyright.
Re: Re: Re:8
Who can say? I doubt they would’ve come up with something great; probably it would’ve been like Stratton Oakmont v. Prodigy and its preceding Compuserve case, and we’d have wanted legislators to patch it up somehow.
But “substantial non-infringing use” is a phrase that American courts came up with, which has featured prominently in copyright cases. That’s reason for “less pessimism”, if not “optimism” per se. Expecting them to have nuked YouTube from orbit seems overly pessimistic.
Re: Re: Re:4
I just saw a story, posted yesterday on Wired, about limiting liability for “A.I.” companies. All of those companies, apparently, decided it was okay to make their tools available without any safe harbor so far. Gun manufacturers also worked for years without any protection. Newspapers printed letters from readers without worrying too much about lawsuits over them.
I imagine lots of businesses would like to have such protections, but it doesn’t seem to be all that critical.
Re: Re: Re:5
There’s also different levels of liability involved. Some of it makes perfect sense: We don’t hold, say, a tool-making company liable for murder if someone uses a hammer made by that company to commit a murder. Some of it doesn’t: We don’t hold gun manufacturers responsible for mass casualty shootings even though the products often used in those shootings are literally designed to murder groups of people in as quick and efficient a way as possible.
We don’t hold websites like YouTube 100% liable for copyright infringement partially because YouTube can’t actually know whether the use of a given copyrighted work is infringing or protected by Fair Use. (Hence the notice-and-takedown system of the DMCA.) But without the DMCA (and likely without Section 230), the chances of YouTube being held liable because it knows about some infringing material on the platform would be much higher. If YouTube could be torn down over copyright infringement, so could any other YouTube-like site. No site would want to try to be the next YouTube if they knew they would just get shut down ASAP by the copyright cartels. The DMCA, for all of its numerous and horrible flaws, was necessary to help keep sites like YouTube from being destroyed.
Re: Re: Re:6
Really? There were several companies vying to be the next Napster, after Napster got shut down. Kazaa, Grokster, and others. Sure, they had some shenanigans with offshore corporate structures, but they did get investors and users.
But I think YouTube had pretty quickly reached a point where it would not be so easily destroyed. There was enough legitimate use that courts would’ve been reluctant to do that (probably; I do share some of your pessimism about it), had they set up any remotely reasonable process for legal complaints. Maybe the courts would’ve said it’s not good enough, and given some guidance for how complaints should be handled, but I doubt they’d have gone “full Napster” and burned it to the ground. There’d’ve been some fine, and some people would’ve complained it wasn’t large enough for what YouTube did, and it would’ve gone on.
Becoming the big central video platform was just too tempting for companies to give up on altogether. The potential profits were huge, everyone knew that, and eventually someone would’ve hit on a formula such that the courts would say “that’s good enough”.
Re: Re: Re:7
Different kind of functionality, different kind of liability. Not a 1:1 comparison here, especially since YouTube wasn’t made with the distinct intent of sharing copyrighted materials and Napster and its immediate successors kind of were.
No. No, it did not. YouTube needed years, not months, to become “too big to fail”; in the time before that, the site could’ve damn well been done in by the copyright cartels. Anime alone could’ve brought down YouTube with the way people were posting whole shows and movies through split-into-parts uploads to fit within the upload limits from back in the day.
You underestimate the copyright cartels and the willingness of the legal system to give them damn near everything they want.
That didn’t happen for a while. I mean, Viacom’s big lawsuit against YouTube was initially filed two years after YouTube had first gone online. While YouTube had become the biggest video site on the Internet, it was by no means “too big to fail” in 2007, even with Google owning the site.
Re: Re: Re:8
Of course; but, as the DMCA safe harbors basically put a stop to the development of jurisprudence around this matter, I’m not sure we can do much better. Anyway, the point’s more that investors are greedy enough to get into even legally-questionable or outright shady stuff, when the potential return is high enough.
Fair enough, but perhaps “too big to fail” is an overly strict standard. “Big enough to have developed substantial non-infringing uses” might suffice.
But let’s go a bit more pessimistic and say YouTube was sued out of existence in its early days. Do you really think we’d have no broadly-similar video-sharing site today? I’m certain that people would’ve kept trying to capture that market. Like with music sites: it was documented that a lot of investors wouldn’t touch music-related technology with a ten-foot pole, due to the RIAA. Still, there were other investors, who kept throwing money until something stuck (Spotify being one example).
Someone would’ve figured out how to make a video site that withstood scrutiny. Like the “pirate taxis” of Uber and Lyft, illegal short-term rentals of Airbnb, and of course the copyright infringement of what’s marketed as A.I. The legal process is often messy, with local rules, circuit splits, and a graveyard of companies sacrified along the way, but tends to work toward sanity (even if never quite achieving it).
Re: Re: Re:
The Chilling effect.
Surely an anonymous takedown notice is on the face of it entirely invalid? If it’s anonymous, how can someone file a counter-notice to the party requesting the takedown?
Cue a completely unknown person/group sending a DMCA claim to get this article taken down in 5… 4… 3…