[Steven_Vellon]

This is the expected result of increased scrutiny over Google. Whether or not to respond to claims of abuse is a tradeoff between limiting false negatives or limiting false positives. When there is an increase in the cost of false negative, both monetary and in terms of status and reputation, then Google is going to optimize to reduce false negatives at the expense of creating more false positives.

This should definitely be part of the conversation when discussing regulating content on Google and other large aggregators and social media platforms. I find that the so-called "techlash" has resulted in more stringent policing of content, but also greater opportunity to abuse reporting functionality. Using copyright claims malicious has been a known practice for years now, at least between content creators (e.g. feuding YouTubers having their fans report rivals). I'm not surprised to see governments using the same tactics.

[dependenttypes]

Google pretty often refuses to respond to counter-dmcas, there is a lawsuit currently over this. In addition recently lumen database stopped publishing the full list of links in the dmca complaints.

[dragonwriter]

> Google pretty often refuses to respond to counter-dmcas,

There is not requirement to respond to DMCA notices or counternotices, doing so merely protects against certain liability you might otherwise have. In the case of counternotices, there generally is no liability to protect against (because providers can usually structure agreements with users to avoid liability for takedowns), and thus no reason to respond.

[jrochkind1]

Hmm, hypothetically if you blanket refuse to respond to coutnernotices, could you lose your protection from liability under DMCA in general? Like even in cases where no notice has been followed? Don't you have to follow the procedures, including responding to counter notices, to be get the protection from liability under DMCA in the first place?

Even if in theory, in practice we know the risk is vanishingly low of actually being held liable for much.

[dragonwriter]

> Hmm, hypothetically if you blanket refuse to respond to coutnernotices, could you lose your protection from liability under DMCA in general?

No, DMCA safe harbor is transactional: if you respond to a particular takedown, you are protected from liability to the issuer of the takedown for the content. If you respond to a particular counternotice, you are protected from liability to that user for taking down the content addressed by the counternotice.

[thebean11]

> no reason to respond.

Sure maybe from a purely cynical profit-motivated standpoint. Even then there's the risk of pissing off your users.

[abakker]

the depressing reality is that most users who are pissed off still don't take their business elsewhere. There are few other choices for similar products that work as well.

[Retric]

Google lost me as a customer. That’s frankly meaningless on it’s own, but add up everyone making the same choice and it can quickly become meaningful.

[Bang2Bay]

there is of course the good old Yahoo!

[ChuckMcM]

I was also going to share this. Google and other places like Twitter and Facebook seem to be willing to pull the trigger instantly on a DMCA and unwilling to put it back up with the C-DMCA even though the same law protects them if they do.

[judge2020]

I'd like to think this is because Google, Twitter, etc. know these media companies have a lot of lobbying connections that could get the DMCA landscape changed to drastically hurt them.

[dependenttypes]

I would rather argue that this is because the media companies are their partners.

[ChuckMcM]

I would lean this way as well. People who are filing counter notices are probably not big buyers of advertising on the platform.

[judge2020]

I think the Lumen full link thing is (at least in part) due to the fact that people were going to the lumen db to find the de-listed links to pirated movies (since the DMCA complaint submitter has already verified that those websites do, in fact, contain the copyrighted movie).

[AnthonyMouse]

> since the DMCA complaint submitter has already verified that those websites do, in fact, contain the copyrighted movie

Well no, actually, that they haven't is exactly the problem. One of the major problems with the DMCA is that it says this:

> 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.

In other words, the claim made "under penalty of perjury" is that you are "authorized to act on behalf of the owner of an exclusive right that is allegedly infringed" and not that the links you allege are infringing actually are infringing.

Meanwhile claims have been known to be filed based on generic searches, e.g. submitting a DMCA claim for every link that shows up in search results for the title of the movie even when it's a generic phrase also used in other contexts, or wasn't quoted so it turned up results where those words (or their synonyms!) appear in any order anywhere on the page. And looking at the DMCA complaint used to allow you to not only find those results that had been illegitimately removed, but also identify that that has happened and then have the ability to object to it in various ways.

[gpm]

As far as I'm aware no court has ever agreed with your interpretation of the wording of the DMCA. Do you have a source on it being correct (which I definitely could have missed) or are you just giving out questionable legal opinions as if they are fact?

The best I can find for someone trying to raise that interpretation in court is when warner bros tried to argue it here [1]. That the motion was denied without prejudice in a brief order here [2], to be re-raised at trial if necessary, and then it looks like the case was settled before it went to trial or the issue was raised again.

[1] https://www.documentcloud.org/documents/835805-184407656-war...

[2] https://www.courtlistener.com/recap/gov.uscourts.flsd.373206...

[dragonwriter]

> As far as I'm aware no court has ever agreed with your interpretation of the wording of the DMCA

As far as I know, no court has every disagreed with that reading, which is simply the plain meaning of each of the words in the statute.

From the day the text was proposed, the fact that the only thing thar a DMCA takedown required to be sworn under penalty of perjury was that the complaining party represented a rights holder has been a frequent point of criticism.

> The best I can find for someone trying to raise that interpretation in court is when warner bros tried to argue it here

No, WB was not arguing that point, they were arguing that a hosts DMCA takedown notice form which purported to require additional certifications to be made under penalty of perjury was invalid because private parties can't just add “under penalty of perjury” without the legal requirements of perjury being satisfied and expect it to have legal effect. The host in that case agreed that what it tried to make under penalty of perjury was beyond what was required in the statute, but argued that the form did, in fact, have effect. IOW, the base requirements of the law weren't in dispute, the effect of a private form as to what was under penalty of perjury was I dispute.

[ikeboy]

>From the day the text was proposed, the fact that the only thing thar a DMCA takedown required to be sworn under penalty of perjury was that the complaining party represented a rights holder has been a frequent point of criticism.

This is wrong. Perjury also applies to having a good faith belief in the infringement.

>The DMCA requires a complainant to declare, under penalty of perjury, that he is authorized to represent the copyright holder, and that he has a good-faith belief that the use is infringing. This requirement is not superfluous. Accusations of alleged infringement have drastic consequences: A user could have content removed, or may have his access terminated entirely. If the content infringes, justice has been done. But if it does not, speech protected under the First Amendment could be removed. We therefore do not require a service provider to start potentially invasive proceedings if the complainant is unwilling to state under penalty of perjury that he is an authorized representative of the copyright owner, and that he has a good-faith belief that the material is unlicensed.

Perfect 10, Inc. v. Ccbill Llc, 488 F. 3d 1102 - Court of Appeals, 9th Circuit 2007

>A copyright owner who submits a takedown notice must include a statement, under penalty of perjury, that she has "a good faith belief that use of the material in the manner complained of is not authorized by the copyright owner." Id. § 512(c)(3)(A)(v)-(vi).

Hughes v. Benjamin, Dist. Court, SD New York 2020

[AnthonyMouse]

Those are literally the words in the statute. I'd be more interested if you could find a case upheld on appeal of someone being actually found guilty of perjury for filing a DMCA claim against non-infringing content. Because if it is never actually enforced then it's a distinction without a difference anyway. In either case fraudulent DMCA notices are rampant and the perpetrators face no consequences.

[ikeboy]

Courts have interpreted the statute to mean that perjury applies to a good faith belief in infringement as well. See two cases cited at https://news.ycombinator.com/item?id=23198236

[gpm]

As far as I'm aware (and again, my knowledge of this field is far from perfect), the justice department has never charged anyone with perjury over a DMCA claim... period.

There have been civil lawsuits, for example Lenz v. Universal Music is famous for establishing that failure to consider fair use when filing a DMCA complaint constitutes misrepresentation under the DMCA. But these don't directly relate to perjury.

[1] https://en.wikipedia.org/wiki/Lenz_v._Universal_Music_Corp.

[dependenttypes]

> people were going to the lumen db to find the de-listed links to pirated movies

Which is a great thing. DMCA should not even be possible against links imo (heck, DMCA should not exist at all).

[ikeboy]

What lawsuit are you referring to?

There's no requirement to respond to counter-dmcas.

[dependenttypes]

> What lawsuit are you referring to?

This one: https://gelbooru.com/index.php?page=forum&s=view&id=4230#1 (warning: this is a porn site)

[ikeboy]

>This is not how the counter DMCA notification system is supposed to work. Upon receipt of a counter DMCA notice, you are required to forward the notice to the original sender, and restore content within 10 days.

This is wrong. The DMCA does not create an affirmative requirement to process counter notices. See https://blog.ericgoldman.org/archives/2020/03/youtuber-loses... for a recent decision affirming that.

Also, I can't find an actual lawsuit filed, just someone talking about how they intend to file a lawsuit.

[derekp7]

That sounds very much like what has been claimed about Amazon handling the French court order to fine them a high dollar amount for each non-essential (as determined by the court) item they deliver. If the cost of something slipping through the cracks is that high, then it is cheaper to not do business where there are potential cracks to slip through.

[jimmaswell]

I'm so glad I don't live in France. What a ridiculous place. People need "non-essential" Amazon deliveries and drivers need jobs now more than ever.

[josefx]

Fun fact: The court order doesn't use the word essential even once, it was introduced by Amazon lawyers and PR. The court order was solely about repeated safety violations and Amazons refusal to reduce operations to a safe level. The court order also gave an explicit whitelist of Amazons own categories for the reduction, not a generic "essential" or "food" classification that Amazon would have to interpret itself.

> People need "non-essential" Amazon deliveries and drivers need jobs now more than ever.

Which is why Amazon quite clearly made the right choice by shutting down the warehouses completely, right?

[MereInterest]

> Amazon handling the French court order to fine them a high dollar amount for each non-essential (as determined by the court) item they deliver

Play stupid games, win stupid prizes. Amazon has shown that they cannot be trusted to determine whether workers or items delivered are essential, and do not take the safety of their workers into account. The cost of behaving in an untrustworthy manner is having everybody else treat Amazon as untrustworthy.

[joshspankit]

Seems to me the onus should fall on the false reporters:

- If the claim is valid, the provider who hosted the content eats the costs as a cost of doing business - If the claim is invalid due to negligence, intent, or abuse of the system, the claimant should face a fine equal to or greater than the true cost of determining that the claim is false.

The courts essentially work this way, the internet should too.

[naruvimama]

Can the claimants not be penalised for false claims?

[ikeboy]

Google sued someone last year for false claims, settled for 25k. https://www.courtlistener.com/docket/16074437/youtube-llc-v-...

I'm involved in a lawsuit against a large company that made false claims of infringement against my business - see https://www.courtlistener.com/docket/16562550/thimes-solutio...

I've been tracking similar cases, and have a database of over a hundred cases of lawsuits over false infringement claims. False copyright claims are often harder to prosecute than trademark or patent cases.

[kevingadd]

In practice it's difficult if not impossible to do anything about it, between providers being unwilling/unable to identify the claimants and the costs involved in taking them to court.

[jandrese]

Especially hard to get the courts involved if the claimants are in other countries.

About the best you could hope for is if a person can be shown to be acting in bad faith they could have all of their takedowns reversed and the material in question flagged so that further takedown notices trigger an audit. In practice this might just mean making a new account for each takedown notice, but it does raise the bar a bit on bad behavior.

I'd also assume big media cartels would be exempt because operating in bad faith is their entire business plan and they're willing to sue to maintain the status quo.

[naruvimama]

A lot of automated copyright claims are made by reasonably sized organisations. Should require a deposit for making automated claims.

Do not have to take them to court, the burden of proof lies with the copyright holder. And sufficient mechanism for counter claims and forfeiting the deposit.

Also can take collective action against all falsely claimed copyright.

The real trouble is something flagged as offensive, once a comment of mine on Quora was taken down. I was given a warning of that being offensive and me being potentially banned. I have no idea to this day what was taken down and why people found it offensive.

The right to contest sadly does not exist, this makes it so easy political or religious groups to silence individuals.

[mullingitover]

Seems like if Google is doing extrajudicial enforcement, they could easily stipulate in their terms of use that false claims can result in a lifetime ban from Google services. People might think twice about making false claims if they (or their entire organization) will lose all access to their Google accounts forever.

[elliekelly]

If Google added that to their terms I would hop in the false copyright claim game ASAP. I want nothing to do with Google and I would be very interested in Google deciding they want nothing to do with me, too.

[ehsankia]

I don't get it, yes false positive removals are an issue worth discussing, but does it really warrant a whole WSJ front page article? Could WSJ make it any more obvious that they have a vendetta against tech companies and are just grasping at whatever hit piece they can get their hands on?

[w-j-w]

Tech companies need to be held accountable for their actions by someone. That's always been journalism's job. As long as they don't print lies, I don't see the issue.

[izacus]

Unfortunately I've seen WSJ (and several other tech media) print outright fabrications and omit whole pieces of stories to drive some agendas. It's hard to keep trusting them now.

[DuskStar]

As long as they don't print lies and don't play favorites, ideally.

[matheusmoreira]

False positives in law enforcement means innocent people are being punished. This isn't something that should ever happen.

[emmelaich]

False positives should never happen and they are inevitable at the same time.

[albertsun]

It's only a tradeoff if you assume that the process must be automated and operate with a minimum of actual human scrutiny or consideration.

Any qualified human review would immediately be able to distinguish and vastly improve both the false negative and false positive rates.

But because these tech companies devalue that kind of work they aren't willing to invest in it.

[polishTar]

A couple of things to consider:

1. There's a real social cost associated with constantly subjecting human reviewers to traumatizing content. It can lead to devastating mental health problems. (some relevant articles: https://www.theverge.com/2019/2/25/18229714/cognizant-facebo..., https://www.telegraph.co.uk/technology/2019/12/16/youtube-mo...)

2. Humans make mistakes as well, especially as volume of content increases and policies get more nuanced.

3. There's way more violating content then you could ever imagine. As just one example, YouTube reported having to remove >1.8 BILLION comments for spam or other policy violations in 2019 (https://transparencyreport.google.com/youtube-policy).

4. The large tech companies spend billions on human moderation as it is. YouTube alone for example has >10,000 full time human reviewers. (https://youtube.googleblog.com/2017/12/expanding-our-work-ag...)

[theonemind]

Regarding 1: I find it unlikely that most DMCA takedown notices contains traumatizing content. (Arguably) most of it has a copyright on it, which means that someone probably wanted to sell it to a sizeable market. The DMCA takedown notices identify URLs, so you don't really need to use the same pool of people that screen for child abuse and various horrors, and they don't get an unfiltered view of every sort of terms of service violation from every platform, just potential copyright violations, a completely separable issue.

[majormajor]

So why not raise the cost of submitting arbitrary user content?

[renewiltord]

Because I, the user, don't want that. I don't have a problem subjecting myself to Content ID in exchange for the ability to just host all sorts of crap.

[tsycho]

Umm, you do realize that this is basically a DDOS attack, and automated tools are a necessity. What needs to happen is a law/legal process where you can counter sue against false DMCA/copyright complaints, with large penalties for indiscriminate and abusive filings. That would put the onus on copyright holders to put in some validation effort before filing.

[bsder]

> When there is an increase in the cost of false negative, both monetary and in terms of status and reputation, then Google is going to optimize to reduce false negatives at the expense of creating more false positives.

The problem is that Google optimizes for cost and doesn't give an iota of damn about accuracy.

Fine their ass when they are legally wrong and suddenly Google will become really good at figuring things out. And, if there is enough cost, they'll even allocate some people to the problem.