YouTube Copyright Transparency Report Shows The Absurd Volume Of Copyright Claims It Gets

from the turn-the-hose-off dept

Any cursory look at Techdirt for stories involving YouTube and copyright issues will give you a very accurate impression of the state of all things copyright for the platform: it's a complete shitshow. You will see all kinds of craziness in those posts: white noise getting hit with a copyright claim, labels claiming copyright on songs in the public domain, and all kinds of issues with automated systems like ContentID causing chaos. That really is a sample platter rather than the whole meal, but it's also worth noting that YouTube knows this is a problem.

To that point, the platform recently put out its first "Copyright Transparency Report" that teases out all kinds of numbers for copyright claims on YouTube videos. As with any statistical report, how you view it is going to come down to how you want to slice and dice the numbers. For instance, it's worth noting that over 99% of the copyright claims YouTube receives comes from ContentID, an automated system. More to the point of this particular post, you will likely also witness copyright enforcement advocates focus on numbers like this from YouTube's own report.

As the report notes, we see low levels of disputes relative to total claims, particularly within tools that use automatic detection. In the first half of 2021, fewer than 1% of all Content ID claims were disputed.

Sounds great, right? Even with some couched language, this looks for all the world like even YouTube is acknowledging that a tiny fraction of the copyright claims are disputed and is therefore a system that is working well. You might even decide to assume that this translates to 99% of copyright claims made at YouTube are accurate and valid.

Except that that is absurd. Vast numbers of people either lack the time, interest, or bravery to dispute a copyright claim like this. Many others may not even realize it's something for which there is a process to take action on such disputes.

And besides all of that, when you dig into the actual raw numbers rather than percentages, it becomes very clear that YouTube does indeed have a copyright enforcement problem.

Over 2.2 million YouTube videos were hit with copyright claims that were later overturned between January and June of this year, according to a new report published by the company today. The Copyright Transparency Report is the first of its kind published by YouTube, which says it will update biannually going forward.

2.2 million claims were overturned in favor of the uploader in a six month period. Let that wash over you for a second and then realize that, while this represents a tiny percentage of the overall claims, it's still a huge number. And when you factor into all of this both that there is zero chance that the actual numbers of invalid copyright claims aren't much higher and the fact that YouTube creators have long complained that this is a massive problem from a process standpoint, well, it again becomes obvious that YouTube has a copyright enforcement problem.

Though mistaken copyright claims are a drop in the bucket on a larger scale, YouTube creators have long complained about how the platform handles claims, saying overly aggressive or unjustified enforcement can lead to lost income. Copyright claims can result in videos being blocked, audio being muted, or ad revenue going back to the rights owner. This new report gives shape to a problem that YouTube itself has acknowledged needs updating.

In 2019, YouTube CEO Susan Wojcicki said in a blog post that the company heard the concerns from creators and that YouTube was “exploring improvements in striking the right balance between copyright owners and creators.”

We most certainly haven't reached that balancing point yet. I anticipate that YouTube is releasing this transparency report as part of a broader and longer term plan to alleviate the problems this is all causing. If not, then all YouTube appears to be doing is telling on itself with this report, which is not the kind of corporate action I would expect.

Meanwhile, thanks to overly restrictive copyright enforcement, YouTube's valuable creator base is left waiting for help.

Filed Under: contentid, copyright, dmca, false claims, overblocking, takedowns, transparency
Companies: google, youtube

Twitter Admits It Messed Up In Suspending Accounts Under Its New Policy, But Policies Like This Will ALWAYS Lead To Overblocking

from the who-didn't-see-that-coming dept

Last week, we called out a questionable move by Twitter to put in place a new policy that banned images or videos that included someone without their permission. The company claimed that this was to prevent harassment and also (for reasons I still don't understand) buried the fact that this was not supposed to apply to public figures or newsworthy media. However, as we pointed out at the time, the standard here was incredibly subjective, and wide open to abuse. Indeed, we pointed out examples in that article of the policy clearly being abused. And even as Fox News talking heads insisted the policy would only be used against conservatives, in actuality, a bunch of alt right/white nationalists/white supremacists immediately saw this as the perfect way to get back at activists who had been calling out their behavior, leading to a mass brigading effort to "report" those activists for taking pictures and videos at white nationalist rallies and events.

In other words, exactly what we and tons of others expected.

And, a few days later, Twitter admitted it messed up the implementation of the policy -- though it doesn't appear to be rolling back the policy itself.

Twitter said Friday that it had mistakenly suspended accounts under a new policy following a flood of “coordinated and malicious reports” targeting anti-extremism researchers and journalists first reported Thursday by The Washington Post.

The company said it had corrected the errors and launched an internal review to ensure that its new rule — which allows someone whose photo or video was tweeted without their consent to request its removal — was “used as intended.”

[....]

In a statement Friday, Twitter spokesman Trenton Kennedy said that the company had been overwhelmed with a “significant amount” of malicious reports and that its “enforcement teams made several errors” in the aftermath.

What I am perplexed about, however, is that I know for a fact that Twitter has a bunch of smart and thoughtful people who work on trust and safety issues and who clearly would know how this policy would play out in practice, and yet for whatever reason the policy was still rolled out the way it was. I don't know if this is because people underestimated how quickly and wildly it would be abused, or if others at the company simply overruled concerns raised by experts, or if it was something else entirely. Either way... it's a weird and surprising misstep for Twitter.

That said, it's also illustrative of a really important point that we've been trying to raise for ages, going back to the DMCA takedown process, and covering all sorts of other policy debates regarding content moderation: if you give people tools to take down some content, they will be abused. Always. That's not to say you should never moderate or never take down content, because that's impossible. There is always going to be some content that sites need to take down, whether for legal purposes or because it's harming the integrity of the site (things like spam or harassment).

But any such policy always opens itself up to abuse and dishonest reporting. And any company that has such policies (i.e., every company with third party content) needs to have a plan in place not just to deal with the abusive/problematic content on the site, but also the abuse of the moderation process to silence voices that shouldn't actually be silenced.

Again, from my interactions with Twitter trust & safety people, I know they know this. And this is part of why I find the rollout of this policy so perplexing.

However, it's also an important lesson for policymakers in various state legislatures, in DC, and around the globe. There is so much effort these days to pressure (or require!) internet companies to remove "bad" content, but almost none of those policy plans take into account the ways in which those policies will be abused to silence reporting, silence marginalized voices, and silence those calling out abuses of power. Twitter's rollout of this policy has been a disaster (and one that could have been prevented) but at the very least it should be a warning to policymakers who seem to think that they can design policy requirements to moderate certain content, without bothering to explore the likelihood that those mandates will be abused to silence important speech.

Filed Under: content moderation, overblocking, photos, private information, private media, videos
Companies: twitter

Upload Filters And The Internet Architecture: What's There To Like?

from the absolutely-nothing dept

In August 2012, YouTube briefly took down a video that had been uploaded by NASA. The video, which depicted a landing on Mars, was caught by YouTube’s Content ID system as a potential copyright infringement case but, like everything else NASA creates, it was in the public domain. Then, in 2016, YouTube’s automated algorithms removed another video, this time a lecture by a Harvard Law professor, which included snippets of various songs ranging from 15 to roughly 40 seconds. Of course, use of copyright for educational purposes is perfectly legal. Examples of unwarranted content takedowns are not limited to only these two. Automated algorithms have been responsible for taking down perfectly legitimate content that relates to marginalized groups, political speech or the mere existence of information that relates to war crimes. 

But, the over-blocking of content through automated filters is only one part of the problem. A few years ago, automated filtering was somewhat limited in popularity, being used by a handful of companies; but, over the years, they have become increasingly the go-to technical tool for policy makers wanting to address any content issue -- whether it is copyrighted or any other form of objectionable content. In particular, in the last few years, Europe has been championing upload filters as a solution for the management of content. Although never explicitly mentioned, upload filters started appearing as early as 2018 in various Commission documents but became a tangible policy tool in 2019 with the promulgation of the Copyright Directive. 

Broadly speaking, upload filters are technology tools that platforms, such as Facebook and YouTube, use to check whether content published by their users falls within any of the categories for objectionable content. They are not new - YouTube’s Content ID system dates back to 2007; they are also not cheap - YouTube’s Content ID has cost a reported $100 million to make. Finally, they are ineffective as machine learning tools will always over-block or under-block content.

But, even with these limitations, upload filters continue to be the preferred option for content policy making. Partly, this is due to the fact that policy makers depend on online platforms to offer technology solutions that can scale and can moderate content en masse. Another reason is that elimination of content and take-downs is perceived to be easier and has an instant effect. In a world where more than 500 hours of content are uploaded hourly on YouTube or 350 million photos are posted daily on Facebook, technology solutions such as upload filters appear more desirable than the alternative of leaving the content up. A third reason is the computer-engineering bias of the industry. What this means is that typically when you build programmed systems, you follow a pretty-much predetermined route: you identify a gap, build something to fill that gap (and, hopefully, in the process make money at it) and, then you iteratively fix bugs in the program as they are uncovered. Notice that in this process, the question of whether the problem is best solved through building a software is never asked. This has been the case with the ‘upload filters’ software.

As online platforms become key infrastructure for users, however, the moderation practices they adopt are not only about content removal. Through such techniques, online platforms undertake a governance function, which must ensure the productive, pro-social and lawful interaction of their users. Governments have depended on platforms carrying out this function for quite some time but, over the past few years, they have become increasingly interested in setting the rules for social network governance. To this end, there seems to be a trend of several new regional and national policies that mandate upload filters for content moderation.

What is at stake? 

The use of upload filters and the legislative efforts to promote them and make them compulsory is having a major effect on Internet infrastructure. One of the core properties of the Internet is that it is based on an open architecture of interoperable and reusable building blocks. In addition to this open architecture, technology building blocks work together collectively to provide services to end users. At the same time, each building block delivers a specific function. All this allows for fast and permissionless innovation everywhere.

User-generated-content platforms are now inserting deep in their networks automated filtering mechanisms to deliver services to their users. Platforms with significant market power have convened a forum called the Global Internet Forum to Counter Terrorism (GIFCT), through which approved participants (but not everyone) collaborate to create shared upload filters. The idea is that these filters are interoperable amongst platforms, which, prima facie, is good for openness and inclusiveness. But, allowing the design choices of filters to be made by a handful of companies turns them into de facto standards bodies. This provides neither inclusivity nor openness. To this end, it is worrisome that some governments appear keen to empower and perhaps anoint this industry consortium as a permanent institution for anyone who accepts content from users and republishes it. In effect, this makes an industry consortium, with its design assumptions, a legally-required and permanent feature of Internet infrastructure.

Convening closed consortiums, like the GIFCT, combined with governments’ urge to make upload filters mandatory can violate some of the most important Internet architecture principles: ultimately, upload filters are not based on collaborative, open, voluntary standards but on closed, proprietary ones, owned by specific companies. Therefore, unlike traditional building blocks, these upload filters end up not being interoperable. Smaller online platforms will need to license them. New entrants may find the barriers to entry too high. This, once again, tilts the scales in favor of large, incumbent market players and disadvantages an innovator with a new approach to these problems.

Moreover, mandating GIFCT tools or any other technology, determines the design assumptions underpinning that upload filter framework. Upload filters function as a sort of panopticon device that is operated by social media companies. But, if the idea is to design a social media system that is inherently resistant to this sort of surveillance, then upload filters are not going to work because the communications are protected from users. In effect, that means that mandating GIFCT tools, further determines what sort of system design is acceptable or not. This makes the regulation invasive because it undermines the "general purpose" nature of the Internet, meaning some purposes get ruled out under this approach.

The current policy objective of upload filters is twofold: regulating content and taming the dominance by certain players. These are legitimate objectives. But, as technology tools, upload filters fail on both counts: not only do they have limitations in moderating content effectively, but they also cement the dominant position of big technology companies. Given the costs of creating such a tool and the requirement for online platforms to have systems that ensure the fast, rigorous and efficient takedown of content, there is a trend emerging where smaller players depend on the systems of bigger ones.

Ultimately, upload filters are imperfect and not even an effective solution to our Internet and social media governance problems. They don’t reduce the risk of recidivism and only eliminate the problems, not their recurrence. Aside from the fact that upload filters cannot solve societal problems, mandated upload filters can adversely affect Internet architecture. Generally, the Internet’s architecture can be impacted by unnecessary technology tools, like deep packet inspection, DNS blocking or upload filters. These tools produce consequences that run counter to the benefits expected by the Internet: they compromise its flexibility and do not allow the Internet to continuously serve a diverse and constantly evolving community of users and applications. Instead, they require significant changes to the networks in order to support their use.

Overall, there is a real risk that upload filters become a permanent feature of the Internet architecture and online dialogue. This is not a society that any of us should want to live in - a society where speech is determined by software that will never be able to grasp the subtlety of  human communication. 

Konstantinos Komaitis is the Senior Director, Policy Strategy at the Internet Society

Farzaneh Badiei is the Director of the Social Media Governance Initiative at Yale Law School. 

Filed Under: article 17, contentid, copyright directive, eu, overblocking, upload filters

Massive Overblocking Hits Hundreds Of UK Sites

from the well,-there's-a-surprise dept

Two years ago, Techdirt reported on a very troubling ruling in the UK courts that BT had to block access to the Usenet service provider Newzbin2. At the time, many feared that this would be the thin end of the wedge, giving copyright companies an easy way to shut down other sites. And with that power, of course, would come the inevitable errors, blocking completely unrelated sites. Just how seriously those mistakes could be is shown by this recent case of massive overblocking, reported here on PC Pro:

Virgin Media and Sky-owned Be Broadband customers found they were unable to access the Radio Times website last week, after the ISPs' anti-piracy filters included the site by accident. Telefonica, which still runs Be Broadband's network said that the overblocking had actually affected around 200 legitimate websites.

Radio Times is a well-known UK TV and radio listings magazine; another major site affected was the citizen science project Zooniverse. As an Open Rights Group (ORG) post explains, the specific Radio Times address that was blocked was radiotimes.com, while www.radiotimes.com continued to function. Here's why:

Many third-party load balanced systems, for example those using Amazon's AWS [Amazon Web Services] infrastructure, are enabled by pointing CNAME records at names controlled by those third-party systems. For example www.example.com may be pointed at loadbalancer.example.net. However, "example.com" usually cannot be directly given a CNAME record (CNAME records cannot be mixed with the other record types needed such as those pointing to nameservers and mailservers). A common approach is to point "example.com" to a server that merely redirects all requests to "www.example.com".

From forum posts we can see that it's this redirection system, in this specific case an A record used for "http-redirection-a.dnsmadeeasy.com", that has been blocked by the ISPs -- probably a court-order-blocked site is also using the service -- making numerous sites unavailable for any request made without the "www" prefix.

As ORG surmised, the problem arose from a UK court decision handed down last month that allowed the Football Association Premier League Limited to block FirstRow Sports, a site for live-streaming sports events. It turned out that the latter used the redirection service http-redirection-a.dnsmadeeasy.com, which was then blocked as part of the court order. Unfortunately, hundreds of other sites, which also used that redirection service, were also blocked as a result.

This is a classic case of overblocking, but on a scale hitherto unseen in the UK. It shows why such Web blocks are very crude instruments, and how easily they can go wrong for quite subtle technical reasons. The problem is that the companies seeking the blocks can make mistakes, but the ISPs implementing the blocks don't want to become responsible for checking that the blocks are correct, and thus implement whatever is sent to them.

A post on the Zooniverse blog explains why this is so problematic:

The fact that the court could issue an order which didn’t see this coming and that the ISPs would act on it without checking that what they were doing was sensible is, in my opinion, extremely worrying. It shows how little power we as operators of a website have -- there are no guarantees that our hard work will travel along the little tubes that make up the internet to make it to your computer, and -- although Virgin were nice in this case -- it's disturbing to think we would have had no redress had they decided to keep blocking us. In the midst of a huge political argument in the UK about filtering content online, it's worth bearing in mind how a simple attempt by a multi-billion pound business to protect its revenue stream ended up, by complete and careless accident, preventing science getting done at the Zooniverse.

Sadly, we can probably expect things to get worse, as copyright companies resort to this approach more often, more mistakes are made, and more overblocks occur with little concern for the damage they cause.

Follow me @glynmoody on Twitter or identi.ca, and on Google+

Filed Under: censorship, copyright, filters, overblocking
Companies: premiere league

Australia Sets New Overblocking Record: Aims For One Site, Takes Down 250,000

from the whoops dept

Overblocking is not a new problem -- over two years ago, Techdirt wrote about an instance where Homeland Security took down 84,000 innocent sites at a stroke -- so you might have thought that those employing this blunt instrument would take a little more care these days. However, things seem to be getting worse, not better. In Australia, the Australian Securities & Investments Commission (ASIC) has just scored a whopper:

The largest number of sites censored when attempting to block one particular site ASIC believed was defrauding Australians was 250,000. Of these, ASIC said about 1000, or 0.4 per cent, were active sites. It said the 249,000 other sites hosted "no substantive content" or offered their domain name up for sale, rather than hosting a fully-fledged active site.

I wonder how the ASIC established that 249,000 had "no substantive content". I can't believe it really checked all of them. And that's the big problem with overblocking: when huge numbers of sites get taken down by mistake, there's no way of telling what just fell off the Web, and what obscure but possibly important information is no longer available.

At least ASIC realizes that its current ham-fisted approach isn't acceptable:

ASIC told senate estimates in its opening statement that it was now examining how it could ensure only a site's specific domain name was blocked and ways it could alert the public to a site being blocked via a pop up page. It was also examining ways such a page could indicate why access was blocked and to whom queries could be made to dispute a block.

That's good, but maybe it would have been better if it had explored those options before shutting down nearly a quarter of a million innocuous sites by mistake.

Follow me @glynmoody on Twitter or identi.ca, and on Google+

Filed Under: australia, copyright, overblocking, takedowns

Indian ISP Penalized For 'Overblocking' In Obeying Court Order To Try To Stop Infringement

from the damned-if-you-do... dept

We've written a couple of times about Indian ISPs overblocking websites in response to vague court orders to try to prevent copyright infringement of a particular film. The responses seemed like massive overkill, such as blocking complete access to certain sites -- including sites like Vimeo -- that have perfectly legitimate reasons for existing. It seems that this overkill may be coming back to haunt some ISPs. Airtel has now been penalized by a Consumer Forum in India for going too far with its blocks. Airtel was ordered to pay one of its customers 20,000 rupees (or about $360) for "deficiency in Internet service, thereby causing mental anguish to the complaintant."

The order said that Airtel misinterpreted the original court order, and couldn't justify its actions by hiding behind it.

“By misinterpreting the Madras High Court order, Airtel blocked entire websites. It is needless to say that the company’s actions amount to deficiency in service as well as unfair trade practice,” said the forum.

Of course, the article goes on to note that there has been some confusion over what was originally asked to be blocked. The anti-piracy company, Copyright Labs, who asked for the block order, claimed that it had only asked for specific infringing URLs to be blocked, but a freedom of information request revealed that it had actually asked for blocking of entire websites.

All in all, this highlights some of the many problems that occur when you give copyright holders the power to order outright censorship. It's good to see some push-back. If other customers in India file similar complaints, perhaps ISPs will think twice before engaging in widespread censorship.

Filed Under: blocking, copyright, india, overblocking
Companies: airtel

How Even Highly-Targeted Censorship Can Lead To Overblocking

from the unintended-consequences dept

As the battle rages over SOPA and PIPA, censorship is very much on people's minds. But there are many different kinds of censorship, operating at different levels of precision. For example, while some forms are crude and inexact, like Homeland Security's shutdown of 84,000 sites, others are highly targeted, and designed to block in a very specific way.

That's the case for the attempted blocking of Newzbin2 in the UK. The judge, Justice Arnold, ordered the ISP BT to take a two-stage approach (possible thanks to the pre-existing "Cleanfeed" system it had set up to block child pornography).

If an IP address matches one of those contained in the blocking list, the request to access the site is routed to a proxy server where the exact URL � the specific address including the directory and filename � is examined. If that matches too, access is denied. The judge specified this technique, rather than simply ordering the IP address to be blocked, for an important reason, which he explained in one of his rulings:

the Studios now accept that the order should refer to IP address re-routing and not IP address blocking. It appears that IP address blocking could lead to "overblocking" of sites or pages that ought not to be blocked.

However, as James Firth notes in a blog post, overblocking is still likely to occur thanks to the combination of two factors:

The first is that Newzbin will - and there's strong evidence they have done already, several times - change their IP address. It is well known that IP addresses have all but run out. Nearly all IP addresses allocated are recycled - they've been in use before.

Pity the website owner who picks up Newzbin's old IP address. Under Arnold, J's BT ruling the new owner of the IP address would have some solace in that the URL would not match, therefore BT customers would still be able to access the website, albeit via a proxy. Re-routing via the proxy may cause some minor problems, but that's a bit of a side issue.

In the case of Sky, [another UK ISP now blocking Newzbin2] unless Sky happen to also use a 2-stage blocking system [like BT] - and my contacts tell me they do not - then whoever picks up the old recycled IP addresses from Newzbin will find themselves blocked.

As Firth points out, one of the many worrying aspects of the Newzbin2 judgment is that there doesn't appear to be any mechanism for removing IP addresses once they have been placed on the blocking list � even if they are no longer used by Newzbin2. Adding new ones, by contrast, is explicitly permitted:

sites "whose sole or predominant purpose is to enable or facilitate access to the Newzbin[2] website" (para 10) can find themselves blocked, again without re-application to the court. If someone creates a website explaining how to work around the block, and this website did very little else but explain how to access Newzbin, it too could be blocked.

That would lead to yet more IP addresses being added to the list, which might then be handed out again to new users as they are recycled from ephemeral sites that pop up and then disappear.

What Firth's analysis shows is how even relatively "sophisticated" censorship systems that aim to minimize damage to other sites can lead to overblocking because not every situation can be foreseen or planned for. Now imagine what SOPA/PIPA will do.

Follow me @glynmoody on Twitter or identi.ca, and on Google+

Filed Under: blocking, censorship, copyright, ip blocking, ip filtering, overblocking, pipa, protect ip, redirects, sopa, uk
Companies: bt, newzbin, sky