Slashdot Mirror
Google Says Almost Every Recent 'Trusted' DMCA Notices Were Bogus (torrentfreak.com)
Reader AmiMoJo writes: In comments submitted to a U.S. Copyright Office consultation, Google has given the DMCA a vote of support, despite widespread abuse. Noting that the law allows for innovation and agreements with content creators, Google says that 99.95% of URLs it was asked to take down last month didn't even exist in its search indexes. "For example, in January 2017, the most prolific submitter submitted notices that Google honored for 16,457,433 URLs. But on further inspection, 16,450,129 (99.97%) of those URLs were not in our search index in the first place."
Why should I check my demands, just carpet bomb them with everything and whatever hits is fine by me.
How about this: Whatever the punishment for not following through with a rightful demand should be meted out for every bogus one made in bad faith.
We used to have a Bill of Rights. Now, with the rights gone, all we have left is the bill.
Google should tweak their algorithm to block takedown requesters who are spamming google with generic requests.
Where are we going and why are we in a handbasket?
DMCA takedown requests for non-existent URLs, especially at a 99.97% invalid rate, should be evidence that the requestor is not properly verifying their DMCA claims and should:
A) Lose their right to continue to submit 'trusted' DMCA takedown requests
B) Be charged under the DMCA for filing false claims.
But we know that will never happen.
If notices had to be signed as accurate by an officer of the company that sent them under threat of perjury, and if it was possible for someone who was affected by a bogus notice to then start the wheels on a process that would see said officer of the company end up doing a few days in jail for that perjury I would think we'd see bogus notices drop to near zero.
But of course that would only happen in a world where the lawmakers actually cared about doing what was right, not what their donors wanted.
That's because Google Knows That Slashdot Editors Don't Know What Their Job Is About
Damn_registrars has no butt-hole. Damn_registrars has no use for a butt-hole.
When you (Google) detect that a medium- or high-volume submitter has "obvious flaws" in more than a very small percentage of its submissions, suspend their "trusted sender" status until the flaws are fixed.
Ditto if more than a small percentage of a medium- or high-volume submitter's requests get overturned or "overturned by default" by an un-challenged counter-notice.
For low-volume submitters, the "kick out" threshold would need to be much higher, something like "5 bad submissions out of the last 10 or 10 bad submissions out of the last 100," along with a "you can't become 'trusted' until we see at least 10 consecutive submissions that aren't challenged and at least 90% of the ones submitted in the last month are okay."
If the legal requirements of the DMCA prevent this, then take "bad submitters" to court and get a court order declaring that you (Google) can ignore requests from that submitter that haven't been approved by the court.
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
I don't have to imagine. I just have to wait for it to get Trump's signature instead.
When were the standards set higher?
Mimetics Inc. Twitter
Nothing. But the URLs are not non-existent, just not indexed by Google, probably due to a robots.txt blocking access to the sites concerned.
You are not alone. This is not normal. None of this is normal.
When I RTFA, it seems like there are 100s of millions of DCMA requests going out each year - 35 million in the "latter half of September 2016"?
Even if "detection" (I don't know how good their detection process is if 99%+ of the claims are not found by Google) filing costs came out to a dollar per then I would think the cost to the industry would be $250+ million.
Looking at Warner Music revenues (the first company I could think of: https://www.statista.com/stati...), I would think that the cost of filing all these claims amounts to around 5% or so of total revenue.
Two things came to mind thinking about this:
1. The music company accountants should be going apeshit over the cost of this program and the damage to the company's bottom line.
2. There's a whole bunch of money to be made creating an application for generating DCMA demands that reduces the number of DCMA requests (and the total cost of the requests) by 10x or more.
Mimetics Inc. Twitter
by mykepredko ( 40154 ) on Thursday February 23, 2017 @02:17PM (#53919283) Homepage
When were the standards set higher?
Before 2002-11-14?
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
Got hit with a bogus violation years ago.
A law firm sent out official looking paperwork en masse, to podunk regional ISPs claiming zomg copyright infringement from various IPs.
Turns out, the law firm was unaffiliated with the rights holder of the songs in question (from an artist I've literally never heard); indeed, the law firm was not even an actual law firm. They did have a nice seedy looking form where you could pay $100 to make your violation just magically go away, however.
My ISP wasn't interested in that, they were merely covering their own ass, but they were happy with my excuse of, "I DUNNO I PLUGGED IN A WIRELESS THINGER MAYBE I WAS HACKED?"
The most prolific submitter needs to be banned from the Internet forever.
When submitting that many million requests, the odds that that high a percentage are legitimate but not in the index is pretty low.
My bet is that the submitter just used an algorithm to generate possible names just in case, rather than actually find real infringers. For example, you know that www.questionablesite.com tends to host some content you don't like, so you do this:
- www.knownquestionablesite.com/bignamemovie.html
- www.knownquestionablesite.com/b1gnam3m0v13.html
- www.knownquestionablesite.com/big_name_movie.html
- www.knownquestionablesite.com/big-name-movie.html
- www.knownquestionablesite.com/big.name.movie.html
- etc
Now you don't know if they were even going to host your movie, but instead of checking, you just spam DMCA takedown requests because they're free and there's no downside if the site doesn't exist.
;^)
Nice to know I've had an impact.
Mimetics Inc. Twitter
Those who are having trouble getting over it are getting in a line, a British line to be exact. They are being queued. :)
Yup it's basically that.
With the additional peculiarity that here, "www.knownquestionablesite.com" will spit a valid page with suggestions, no matter what you throw as a name afterward (even if "big.name.move.html" isn't in their database, it would still give a list of not necessarily related download links).
So they are not exactly issuing DMCA about links that don't even exist (these URLs do not return 404s), only DMCA about links that are not in google database (random links that elict a random answer from the website).
The claim is borderline bogus because, as mentioned, the website return random non related download suggestions. So the website is not necessarily infringing on the DMCA submitter's IP. On the other hand, as the result page is random, Google can't prove that the submitter didn't get their IP showing by random chance on the result page on the precise occurence when they tested the URLs about which they decided to file the complain.
So currently Google is deciding to accept the submissions. But that could easily get changed in the future.
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
Legally defined bad faith is hard to prove. We would need some metric, like number or proportion of bogus requests where bad faith is legally presumed. Of course, if a reasonable person hearing/viewing the target of the takedown knows it isn't the complainant's property, that too should be presumed to be bad faith..
Someone charges by the numbers for those take down notices. Someone at the media conglomerates got fleeced.
The only thing worse than a Democrat is a Republican.
The question I have is: what makes a notice "bogus"? It could mean at least any of the following: the URL never returned any useful content (returns a 404 or similar); the URL used to return content, but does not currently; the domain is not responding; the URL is not indexed by Google; the URL returns content, but none that contains a whiff of the copyrighted material; or other possibilities.
If the only thing "bogus" is that the URL is not indexed by Google (but does impermissibly contain copyrighted material), then this shouldn't be a problem. Google quickly scans the URL, sees that it's not indexed, and tosses the request. There should be no need for human intervention. If they wanted to, Google could even add the URL to a list of DMCAed URLs to prevent possible indexing in the future. All automated responses, very cheap to implement.
As for why copyright holders would do this, it's not difficult to see. Say you find a URL infringing your material. You don't know who indexes it. You can check Google, Yahoo, Bing, etc - but they are constantly updating their results, and it may not show up at the time you search. Especially if pirate sites appear and disappear quickly. So you send DMCA notices to every major search engine so the ones which index it stop and the ones which haven't indexed it don't pick it up later. Multiply this by 1,000 and you see why it's infeasible to verify that every takedown notice is currently active in Google's index.
If you want to talk about whether the DMCA process itself is reasonable, that's another conversation. But given the current system and the definition of "bogus" implied by the summary, this approach doesn't look unreasonable.
If infringing content is usually hidden behind robots.txt blocks, then this makes a fair amount of sense.
I suspect it would mean that in order to find the infringing content, the people making the takedown requests are likely violating the robots.txt. I doubt that's enforceable, unfortunately.
When 99+ percent of your claims are bogus, it's pretty much a given that you're not acting on good faith.
Just leave the bad faith out if it bothers you. You make a bogus claim, you pay.
We used to have a Bill of Rights. Now, with the rights gone, all we have left is the bill.
Sure, to you and me it's a given. But it needs to be written into law so we don't get bizarro decisions in court.
That's probably why Google is publicizing this. To point out that the DMCA badly needs a disincentive for filing false takedown claims. If only 0.05% of claims are even factually correct (not even considering if they're legally valid), that's a huge problem with the law.
Some sites generate pages for every URL you try to access on their domain. site.com/movie-name.html is auto-generated when someone tries to access it and not stored anywhere. Google's bot recognizes this and doesn't bother listing the page, but the dumb DMCA bots are apparently not so smart.
That's why sites with zero actual content can attract hundreds of millions of DMCA notices. I do wonder if they are honeypots designed to waste the bot's time or testbeds for anti-DMCA-bot technologies.
const int one = 65536; (Silvermoon, Texture.cs)
SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
That's why our judges get some leeway when it comes to interpreting laws. Else you get people who search for loopholes and get away with it.
We used to have a Bill of Rights. Now, with the rights gone, all we have left is the bill.