438 points by raleighm 10 months ago
The major issue in this is not the bot/automated process that initially flagged the video (although I don't agree with that either).
The primary issue is that the appeal process was rejected. This shows that either all appeals are rejected outright or the person who was physically responsible for checking it didn't read the appeal or view the video.
I'm not sure how anyone could reasonably assume "this is me personally performing a 300yo song" - either by reading the appeal notes or viewing the video could assume that it breached copyright.
This is why there needs to be damages / fines for all incorrect appeal rejections. Because if there's no penalty for getting it wrong, there is no motivation for a company to actually do anything about it.
If it costs them $1000 - $10,000 for getting it wrong, it will be more cost effective for them to pay someone $20/h to double check before rejecting an appeal.
Somebody presented the idea here on HN the other day about $1 fines - both for individuals and for the music copyright abusers.
The nice thing about this IMO is 1$ is a small amount for me and you, but for a big corporation like sony this will cost them thousands every time it gets invoked (because legal will probably get involved).
Of course it needs to be slightly more advanced I guess but it would be a good start.
At most one guy from finance will be involved to rubber stamp the budget for the quarter.
I don't think so. The idea that they should pay anything back will probably be very hard for them to wrap their head around. ;-)
No, it won’t. Corporations understand the phrase “cost of doing business.” If you want fines to deter corporate wrongdoers, they have to be substantial relative to the company’s profits or revenue.
This is probably a good idea, but it's not really implementable. Google isn't a law enforcement entity, so they are probably not in a position to fine anyone.
And IANAL, but charging for claims is probably not allowed by DMCA.
Google can ban youtubers for breaching a threshold of community guidelines.
Apply the same to process to incorrect flagging of copyrighted content. Break the rules, prevent Sony from uploading for a day or two (or some other functionality).
This will never be implemented because there is a symbiosis between a mega content creator and an advertising company.
Sanctioning a producer like Sony would be a sanction against google itself.
I'm afraid the little people will have to suck it up.
Or, decentralize, and stop being controlled.
I agree. I'm going to give Mastodon a try.
Private groups can levy fines. They can raise the costs of ads for these publishers on their platform, take a portion of the revenue from their content, or simply cease doing business with them until it's paid.
For example, late fees on unreturned video cassettes.
ContentID has absolutely nothing to do with DMCA takedown notices, especially in a legal sense.
They might use a computer with possibly AI to handle it. So the procedure must involve something like a CAPTCHA or human-human-interaction to be effective.
All divisions, subsidiaries and parent organisations lose all their copyrights for any false claim.
Or a cost of $1000 or so per flag, refundable upon the copyright infringement being valid.
There is a large amount of piracy which gets legitimately flagged. It's unfair to ask copyright holders to perpetually "loan" millions of dollars to hosters, which is what this scheme would effectively enforce.
"It is better 100 guilty Persons should escape than that one innocent Person should suffer". That's been a theoretical basis of the criminal justice system (or at least something it claims to aspire towards).
The argument that lots of crime happens and it's difficult to make sure that efforts to police the law only target the guilty doesn't hold up under that. We're not meant to bias towards punishing a few innocent people because of that difficulty.
This is disproportionately unfair to small business in the space, tho.
Ok, how about a $1000 to remove the the content immediately, refundable upon investigation, otherwise the content stays live until the investigation finishes?
Then perhaps make the charge per block of reports?
How large a block should be could be a difficult thing to judge. Perhaps a bock could just be all the reports made in one day. Perhaps let the reporter chose: if anything in a given block of reports is rejected then they all are and need reporting again, this would make the reporter group the definite infringements together but be more cautious with the more spurious/speculative reports. Though a concern with any system like that is implementation and UI/API complexities.
Can they not file a civil suit to recover damages?
It'd cost more than you'd likely recover. Even a video with a lot (1M+) of hits only generates a few hundred to a few thousand dollars of revenue (depends on how many of those users skip the ads basically).
Might be worth trying to put together a class action suit for all of the people that this keeps happening to... but again, the recovery would likely go primarily to the legal firm, though it might be enough to get Google to wake up and do something about their auto-rejecting appeals system.
Small claims, filed in your local county. Sony isn’t going to send a $500/hr attorney to fight a $5000 suit.
I won a case against American Airlines doing that (a baggage related issue.)
Can you file against Sony in your local county when the problem happened via an international company like Facebook or Google, rather than at your local airport? I have no idea, but I'd like it if there really were a simple, and cost-effective way for the little guy to throw some rocks at Goliath now and then in egregious cases like this.
Sony does business everywhere, and if it's flagging the copyright in your country I assume it is possible to sue the local company.
"county" isn't the same word as "country". In the US, you bring small claims suits in a county court...county is very localized, and there are thousands of them in the US. Usually the terms of service of video hosting sites will require lawsuits to be brought in the home county of the company (probably somewhere in Delaware or California for Facebook and Google), and they're likely the offending party here (since Sony files through Facebook and Google, and those companies are the ones who're imposing a terms of service that prohibits uploading "copyrighted" materials), so it is not at all clear one could bring a small claims suit against Sony in your local county. If fact, it seems unlikely, based on my understanding of the law.
I wish it were possible, but I don't know that a county court would consider itself to have jurisdiction in such a case.
Yes, you live there, so the violation happened in your country.
If you live in an "insignificant" country it is possible Sony as no presence and so your win won't get you anything. However in the US or most of Europe Sony has enough presence that some part of them will have to pay.
I said "county" not country; you're having a different conversation. We're discussing small claims court at the county level. You don't sue someone in "United States court", you sue them in a county court, usually in the county where the matter being sued over happened, but sometimes the appropriate jurisdiction is the home county of the person or company you're suing (if it's not the same as where the matter took place).
The best explanation for this I could find is here: https://www.nolo.com/legal-encyclopedia/free-books/small-cla...
But, it doesn't really answer the question. The agreement between Facebook or YouTube and the uploader definitely requires suits be brought in the jurisdiction of the company (so, Delaware, probably, but maybe California). Sony is using Facebook or Google to do their dirty work in these cases...so, who can you sue and where? I don't think it's at all clear that Sony could be sued in a local small claims court for a matter like this, though I would like for them to be.
Sony might if that suit could set unfavorable precedent, no?
Not sure how much precedents matter in small claims. It usually boils down to how the judge feels about the issue.
Small claims judges in the US are generally supposed to follow precedent, but they don't make precedent. Regardless, in most cases a small claims case can be appealed to a regular circuit court and receive a de novo (completely new) review. (of course, for a corporation, that means hiring an attorney...)
Might be worth trying to put together a class action
But first check the terms of service to make sure you didn't agree to mandatory arbitration and waive the right to class action.
Exactly usually to use these services you basically wabe away any and all rights.
"...And if a takedown notice is wrong, a counter-notice can be sent by someone explaining that they own the work or that it’s not infringement..."
I think "if a takedown notice is wrong" - counternotice need to be sent back with a mandatory bill to pay to compensate for effort of receiving party to prove that takedown notice is invalid.
This will curb down this BS practice of frivolous opportunistic notices.
That's true of the American legal system as a whole. It generally seems to ignore the cost to the victim of dealing with the legal system...
There is no good solution to that problem. Loser pays is even worse as you have to risk the loss if you incorrectly lose in court.
Loser or taxpayer pays, based on the judgement of the court?
That means I as a taxpayer lose when someone screws up.
Some of the time, yeah. It seems a lot better for all of us to front theses costs of government operation rather than foisting all of that cost onto people who were unlucky enough to be wronged or wrongly sued, by no fault of their own.
Cheaper than the company hosting dealing with tons of lawsuits unfortunately.
This isn't a legal takedown notice though, it's facebook's own system to keep the record producers happy. They can do this with impunity because facebook/google's system is designed to help them takedown without having to go through legal notices, which i suspect would have to be filed by a human atleast.
And an additional layer, if the victim has the time/resources to countersue, and can manage to prove malice or negligence behind the takedown, the punitive fine should be absolutely massive.
The best part is that the DMCA theoretically has provisions to protect you against fraudulent claims, but no platform holder implements them. You can get absolutely blanketed by fraudulent DMCA notices and you have no recourse because companies like Google refuse to provide you any way to sue the person who's violating the law to harass you, and the companies don't make any attempt to filter out fraudulent notices.
Some random kid in Iowa could cost you thousands of dollars in daily revenue for weeks (while you wait for your dmca'd content to come back up) and you have no recourse because the law doesn't have any teeth.
Thats because technically, a youtube takedown is not a DMCA request. Instead it is a request closely modeled on DMCA takedown requests.
This applies to other Google services without content ID, like their extension store and app store
And then you enter the realm of private settlements where the victim is given money to drop the suit and to never talk about what happened ever again.
Technical solutions to this seem to be hindered by issues that affect other secure systems, such as: how do you know who first published something (trusted/reliable time stamping servers?), and how do you know it’s the same content?
As such, I suspect the best system is “hire a bunch of humans to check”. And that’s expensive, which is fine because it should be. These companies try to completely decimate online publications just because it might infringe on their precious properties, and that needs to be a massive cost to deter frivolous takedown notices.
It’s kind of amazing that sites so rapidly reached the point of just insta-killing potentially infringing content with no evidence. That has surely crushed some up-and-coming artists who probably said “screw it” after having their platform taken out from under them by mistake.
The key take away that I get from this article:
"How many more ways do we need to say that copyright bots and filters don’t work? That mandating them, as the European Union is poised to do, is dangerous and shortsighted?"
That is assuming that false positives for small artists are something that the big EU publishers worry about. They don't. Destroying the public domain and indie works is acceptable collateral damage for the politicians and a desired secondary effect for the publishers. The copyright filters will work exactly as designed.
I'm not assuming that the publishers worry about this -- I'm saying we should.
I don't understand why companies aren't being charged with perjury for putting false information on a takedown form. Prosecute a few of them, and the problem will fix itself.
Maybe not perjury, because DCMA takedown requests aren't sworn testimony. But this case isn't even about DCMA. It's just Facebooks takedown service to avoid DCMA invocation.
I'm not even sure who the victim would sue here. Facebook or Sony? Both, I guess, and let them sort it out ;) And yes, maybe in small claims court.
The law states you have to include a statement like this in the letter:
Under penalty of perjury I certify that the information contained in the notification is both true and accurate, and I have the authority to act on behalf of the owner of the copyright(s) involved.
If you don’t it should be dismissed outright.
Thanks. Didn't know that. But then, how does this apply to takedowns handled by Facebook etc? From what I've read, they aren't actually DMCA takedowns.
Isn't it kind of crazy how such a huge company can be so outlandish that it requires the general public and a musician with enough following to question such absurdity to force a human response?
That's reminiscent of issues re Google nuking accounts.
What gets me about when giant corporations such as Sony or others absurdly overreach like this is that they never publicly say sorry.
Just saying “sorry the process got confused, of course we don’t own the works of Bach” would go a long way.
No, its not even a single bit outlandish, it is the normal procedure how cooperations operate that are too big to be affected by markets or laws. Google, and Facebook are big enough to have a virtual monopoly in their respective domains, so the logic of markets doesn't apply to them anymore. They are also big enough to brazenly ignore certain laws, soley based on the size of their legal departments -- they might be successfully sued in some instances, but it doesn't really matter in the big picture, as long as the average joe is deterred from taking action ("they have so much money for lawyers, you don't have a chance").
These conditions make actions like this not only "not outlandish" but outright reasonable (as seen from the logic of the cooperation). If the system that encourages this behavior doesn't change, the cooperations certainly won't act any different.
This seems perverse in the context of the copyright legislation being discussed in the EU today.
On one hand, rights holders can indiscriminately send out takedown notices to any they suspect of infringing, and the burden is then placed on the poster to prove that it's OK. On the other hand, sites will be required to police all posts, just in case something infringing is posted.
So #1 - it's OK for the rights holders' bots to be inaccurate, but sites will be punished if theirs are.
and #2 - all responsibilities accrue away from the rightsholders. They get all the rights, but have none of the responsibility or burden for (accurately) watching over those rights.
As EFF notes, this is great for him, but horrible generally, in that chances of success against jerks will likely be low.
Also, I wonder about speculation that Sony might have owned copyright for the score that he played.
I don't think I want to live with a legal system that claims that my performance (as a human player) of someone else's sheet music is somehow bound by copyright. I draw that line at recorded performances, wherein, say, the orchestra might have reproduction rights to their recording of a public domain work.
I think if the law is any more strict, things like covers wouldn't be able to exist legally, and looooots of genuinely entertaining small-time youtube entertainers that develop their talents by singing or playing covers would suddenly be under legal fire. Even a complete novice can sing along to something on the radio, so trying to claim that they shouldn't be allowed to do so fails my smell test.
Covers exist because of intra-music-industry rules: so-called "mechanical licensing". Arrangements and original compositions are very much subject to copyright...so this is the world you're living in, I'm afraid.
OK, but does the copyright for a score or an arrangement apply to just the thing itself (as in a file or printout) or also to music performed using that score or arrangement?
That's complicated, but here's my layman's understanding.
If you write down an original song in a score, then you have copyright for that song, including performances of that song. This is actually true for any time an original song is first placed in a fixed medium, so if you record a song but don't write it down, then the rules are the same.
If you make a novel arrangement for a public domain song, then you have copyright for that specific arrangement, including performances of that song.
If you transcribe a public domain song, then you may have copyright for that transcription (so if someone were to photocopy it, it may infringe), but you do not get any copyright to songs performed from that transcription.
Radio play of a song pays a fee to the owner of the writer's copyright, but not the performer of a song, so David Paich & Jeff Porcaro (or whomever paid them for ownership of the rights) are getting money every time Wheezer's cover of "Africa" plays on the radio, while Wheezer does not get money.
The performance using it is a derivative work for which you'd need a license to create (in UK) or perform, or publish.
> I don't think I want to live with a legal system that claims that my performance (as a human player) of someone else's sheet music is somehow bound by copyright.
It's going to be hard to find a place where you'd want to live then. I agree that this is a thing that should not exist but it does and the practice is very wide-spread. For an encore: There was a case (I really wished I could find a reference) iirc where a band had sold the rights to a piece and they were then required to pay royalties on a performance of their own creation.
That's distinguishable, I think. In your example, I gather that the band sold rights to a piece, and not just to a particular score or arrangement of the piece.
How would you distinguish a performance based on the public-domain Bach-Gesselshaft Ausgabe from one from a different score? Also I don't think Sony prints editions of classical texts.
That came up in a recent thread about this issue. As I recall, someone claimed that expert listeners can tell what score was used.
Experts supposedly could tell expensive wines from table wines... this doesn't pass muster for me anymore either.
<rant>I'm getting to the crotchety old man stance of - drop copyrights, drop patents, everyone gets paid for [a live performance|first sale of a physical object] and screw all this secondary income bullshit. None of it makes sense and it leaves too many openings for pirates/scumbags/greed to suck money out of the public. </rant>
There's certainly a strong argument for rolling back all the extensions and expansions for patents, copyrights and trademarks, which have been rammed through over the past few decades, by exploitative rights holders.
The alternative is just to ignore it all, and use adequate OPSEC to avoid prosecution. If the system is so totally dominated by monied interests that there's no hope for relief, what else can we do? Just declare independence, and let them pound sand. China and Russia have at times done that, more or less. And we can create our own overlay Internet that can't be coerced or taken down. Or at least, can't be taken down without unacceptable consequences for the commercial Internet. If it comes down to it, overlay routers can spread as malware, and use covert channels, hidden in HD video etc.
I've actually been really curious, does anyone know what kind of algorithm they use to detect similarities?
From what I've read of the tech behind Shazam/SoundHound, I was under the impression that the kind of audio fingerprints produced and indexed relied on the waveform/FFT and rhythmic input being identical, and that a band doing a really accurate cover would never trigger a match because it would be totally different in a million tiny details, the details used for the actual fingerprinting.
But clearly that's not the case here. No two pianists play perfectly alike, so what kind of algorithm is robust enough to match the underlying composition as opposed to the execution of it?
There's got to be some fudge factor in algorithms used by e.g. Shazam -- an FFT of a microphone recording is never going to be strictly identical to the FFT of the source material.
Shazam predates the recent big Deep learning movement iirc. I can conjecture but I feel tech for "understanding" the music has come a long way since Shazam first started. Esp when you see YouTube labelling music in personal recordings.
EDIT: also see  for a semi high profile case of a busker getting banned from FB for doing a cover
Last year I received a YouTube copyright claim against my use of an extract from the Open Goldberg Variations:
which I presume was generated by an automated process that couldn't accurately distinguish between performances.
This just demonstrates how absurd automatic filtering is. It simply equals blind censorship. Usage or presumption of guilt is a common pattern in copyright abuse, from such takedown methods to DRM and other copyright inspired censorship garbage.
Earlier thread on this: https://news.ycombinator.com/item?id=17923138
So we went from small copyright violators being able to steal without repercussion to large copyright violators being able to steal without repercussions.
Social media went from one extreme to another. Now the big corporations can exploit social media to rob and steal.
This is exactly what it is. Theft by sony and other large companies. Since social media doesn't punish sony and large corporations for abusing the copyright claim system, they will abuse it.
Small copyright violators get suspended if they continue to abuse the system. Each social media platform has a threshold after which your account is banned. Why are corporations not held to the same standard?
If sony knowingly and willingly makes false claims X number of times, then their account should be banned and they should be prevented from making claims. Otherwise, they will claim everything since there are no consequences.
It's sad how social media went from user/people centric to corporate centric in such a short time. Youtube is no longer "you"tube, it's "CNN/HBO/SNL/Late Night Show"tube now.
Everything will devolve to control by rich and powerful unless the public acts to check it.
Is there any reason this can't be done in reverse? (IANAL) Perhaps a small orchestra can put a bunch of stuff online, with copyright of their performance, and then issue takedown notices on all of Sony's versions.
Getting access to the automated process on a platform requires that you are a large player. If you use the smaller system to issue notices, there is no plausible way you aren't acting in bad faith.
That plausible deniability offered by the automated system probably gives the large players sufficient cover.
Interesting. So what does happen with dueling takedown requests? Do first movers always have the advantage?
I'm not sure (not a lawyer). In theory, both request should be squashed, and I think there's a serious penalty for knowingly issuing a bullshit request, but i've never heard of the big guys being sanctioned, perhaps they lean on the excuse that their automated tools generate false positives. So you need to put up the content, and then point crappy automated tools at the big boys, maybe. I'm not gonna do it, but if someone who knew what they were doing managed to generate a ton of takedowns at the large providers, in a legally safe manner, it might at least prove a point and generate some press.
But will they give Bach all the lost revenue due to takedowns?
Is this sarcasm?
What if I told you...there was an app on the market...
Bach died in 1750.
Thanks, that made my day!
Sony must be brought sternly to account for this fiasco. The penalty for this type of mistake should be costly enough that it will never happen again. If the United States will not update the DMCA to sharply penalize false copyright claims, perhaps North Korea should step in and address the situation.
Hopefully the fine will teach them a lesson?
Who's getting fined, and by whom?
No one. That's the joke.
Not sure why the link to Sony disagreeing that they don't own copyright to Bach links to a tweet by Jeff Bezos.