Senator Tillis Plans Major Copyright Overhaul: Recognizes Legit Problems, But Current Solutions Are Lacking

from the pandora's-box? dept

We've criticized Senator Thom Tillis for his patent and copyright reform ideas that seemed to take a strong "maximalist" line in its approach. He's also taken the Hollywood line on things like the Internet Archive that was troubling. He's been promising a copyright reform bill for a while, and many thought it wasn't going to matter much, since polls indicated that he was very, very likely to lose his re-election campaign in North Carolina (even after his opponent's bizarrely chaste sex scandal made the news). However, Tillis pulled out a surprise victory, and that means that his plan to reform copyright is something worth watching.

Earlier this month, he sent around a a letter to "stakeholders" asking for thoughts on a wide variety of issues related to the DMCA -- which actually suggests that he might be open to some good changes, and not just the terribly awful ones we've been hearing from Hollywood over the last few years. There are still many horrific ideas in the letter, but the fact that so many broad ideas are in there at least (hopefully?) suggests that Tillis' office is willing to look at the entire DMCA structure, and not just pasting on another favor to Hollywood (which had been Congress's traditional approach to copyright law for decades). The letter even states this upfront:

Rather, than tinker around the edges of existing provisions, I believe Congress should reform copyright law’s framework to better encourage the creation of copyrightable works and to protect users and consumers making lawful uses of copyrighted goods and software-enabled products, respectively.

The very fact that it states upfront that the law should protect users and consumers is a surprisingly good start. Many of the questions in the letter could lead to good, bad (or neutral) results depending on how the eventual bill answers the question. One good aspect is that it appears that Tillis is skeptical of DMCA 1201, which is the "anti-circumvention" or "digital locks" provision in the DMCA that says circumventing "technological measures" to lock down content is infringement, even if the circumvention is not done to infringe on anyone's copyright. This has been a mess in practice, and has resulted in fairly widespread abuse of the law to lockdown products and effectively take away ownership rights from buyers. It's also been hell on security researchers who risk tremendous liability for just doing their jobs.

1201 has this incredibly messy "triennial review" process by which the Copyright Office and Librarian of Congress decide to "exempt" certain applied for use-cases from 1201's rules for the next three years. But as we've seen time and time again this whole process is very sketchy and can even lead to rights being taken away at a later date. Also, frankly, if your law needs a review with new exemptions every 3 years, that seems to suggest the law is the problem. Tillis' questions seem to recognize some of the problems with 1201 and the triennial review process:

Section 1201 currently allows for temporary exemptions to be granted from the circumvention prohibition, but those exemptions do not extend to third-party assistance. This means that when the Librarian of Congress grants an exemption for circumvention of technological protection measures (TPMs) over software for a tractor to allow for repair, the tractor owner must perform the software repair themselves. The Copyright Office has recommended amending the statute to grant the Librarian authority to adopt temporary exemptions permitting third-party assistance “at the direction of” an intended user, and this may be the right way to address this problem. Do you agree with the Copyright Office? If so, how should this provision be drafted to avoid unintended consequences, and to what extent is the Unlocking Consumer Choice and Wireless Competition Act a helpful model? If not, please explain why you do not agree and provide specific recommendations as to how you think this problem should be addressed?

The Copyright Office has recommended revising some of the permanent exemptions so that they are better tailored to the types of uses sought today. In particular, the exemptions for security testing and encryption research should be revised to expand the types of activities permitted, ease the requirements to seek authorization from the owner of the relevant system or technology, and eliminate or clarify the multifactor tests for eligibility. What thoughts do you have about revising these existing permanent exemptions, and how would you recommend that be done?

Congress should adopt new permanent exemptions for noninfringing activities that have repeatedly received exemptions in recent triennial rulemakings, or where there is a particularly broad-based need, including to enable blind or visually impaired persons to utilize assistive technologies and to allow diagnosis, repair, or maintenance of a computer program, including to circumvent obsolete access controls. What other temporary exemptions should be made permanent?

There are various ways that the triennial rulemaking process could be streamlined to be more efficient and so that section 1201 better accounts for user concerns. These include establishing presumptive renewal of exemptions adopted in the previous rulemaking cycle, shifting the burden to those who want to oppose an exemption from the previous rulemaking, and authorizing the Librarian, upon recommendation of the Register, to make permanent a temporary exemption that has been renewed twice without opposition and without modification. How ought section 1201 be revised to reflect the stakeholder desire for a less burdensome triennial rulemaking process and consumer interests, and what other means should be adopted to make the rulemaking process more efficient?

It's good that Tillis recognizes the problems of the triennial review, though I wish that the office was more open to the fact that all of 1201 is outdated and not useful (and open to widespread abuse). Rather than just looking at fixing the triennial review process, it would be even nicer if Tillis was open to just dumping 1201 altogether. He noted in the opening that he's not interested in tinkering around the edges -- but many of these questions indicate that is what's happening. A real review would question the entire premise of 1201 itself. Is it actually needed? Is it useful? Is it promoting the progress of knowledge? Or is it simply a tool for taking away user rights and harming innovation?

Most of the rest of the questions focus on the more well known parts of the DMCA, Section 512, which is the "notice-and-takedown" part of the DMCA, or the "safe harbor" part. 512 has always been problematic, and there's a long history of it being abused for massive censorship. I still think there's a decently strong argument that 512 itself violates the 1st Amendment, in that it uses state pressure to remove 1st Amendment protected content.

Tillis seems to recognize that 512 is outdated, but it's a very mixed bag in how outdated it is. Some of his suggestions are very much needed, but a few suggest he's considering making 512 even worse and even more censorial. Again, starting with the positive, the letter recognizes that 512's classifications of service providers is weird and no longer in-step with reality. Currently it has different rules for conduits, caching services, hosting services, and location tools. But in many 512 cases, it's not always clear which category sites fit into, and it's not always clear how the different rules apply to different types of services. Clarification on that would certainly be useful -- though, again, what that clarification looks like will matter a lot.

OSPs eligible for the safe harbor under section 512 are divided into four categories (conduits, caching services, hosting services, and web location tools) that can be both under-inclusive and over-inclusive. First, what types of OSPs should be covered to account for technological advances and business practice changes that have occurred during the past twenty-two years? Second, how should the categories be revised to better cover the types of OSPs that need—rather than just appreciate—the safe harbor’s benefit? Among the possibilities would be to either increase the number of statutory categories to more explicitly cover specific types of service providers or to reduce the number of statutory categories, possibly to only one, and delegate authority to the Copyright Office to identify, by regulation, the covered types of service providers. If Congress were to take the latter approach, would this raise concerns about such authority being delegated to a non-presidentially-appointed Register?

Of course, neither suggested approach seems likely to be that useful. Simply increasing the categories is going to lead to the same problem down the road (and could impact future innovations as companies "choose" which category to go into). And having the Copyright Office pick and choose would also be a disaster (and arguably unconstitutional). A better approach would be to go in the other direction, and stop trying to fully distinguish types of services but to focus on types of uses. Indeed, the best fix to the 512 safe harbors is actually to align it with Section 230's immunities. That is, rather than put in place a series of specific conditions you have to meet to "earn" the protections (which has been a mess in practice), just set up an immunity that makes it clear that service providers are not liable for the actions of their users. Simplify the situation and let innovation bloom.

Another potentially good thing in the questions is that Tillis recognizes the counternotice situation for many users is not good, and is leading to non-infringing works that are being taken down remaining down because users do not wish to poke the bear. A counternotice under existing law is effectively daring the copyright holder to sue. And even if you believe strongly that your use is non-infringing, going through a federal lawsuit to prove that would bankrupt many people. That's correct, though Tillis' proposed solution of a "small claims" court creates many other problems that we've discussed in the past. There needs to be a better way, but the small claims proposals we've seen to date will lead to more trolling.

The questions also recognize problems in automated takedowns, and potential privacy issues with notice-and-counternotice situations (in which abusive individuals file bogus notices in order to try to dox someone by receiving their counternotice, including contact info). Again, it's good to recognize these problems, even if the suggested solutions leave much to be desired.

It is clear from the record established across my hearings that one major shortcoming of section 512 is that users who have had their content removed may decide to not file a counter-notice because they fear subjecting themselves to federal litigation if the copyright owner objects to the putback. At the same time, the requirement that a copyright owner pursue federal litigation to keep a user from having content put back up following a counter-notice is a heavy burden. Congress might consider improving dispute resolution by directing disputes between notice and counter-notice filers to a small claims court rather than federal court. What is the best way to accomplish this? Would the copyright small claims court as envisioned by the CASE Act be the proper forum? If not, how should such a tribunal be designed? Related, what should be the time period for putbacks? There is broad agreement that the current 10-14 day window works poorly for both copyright owners and users. How would you amend this?

More generally, the notice- and counter-notice sending process have many shortcomings. These could be improved by clarifying when automation is appropriate and that OSPs cannot erect requirements beyond those in section 512(c)(3); by authorizing the Copyright Office to develop standardized web forms for notices and counter-notices and to set regulations for the communications that OSPs must deliver to a user when their content is taken down or had access disabled (including offering information about the fair use doctrine as codified in section 107 and as illustrated in the Copyright Office’s Fair Use Index); and by increasing privacy protections for notice and counter-notice senders by masking certain personally identifiable information, including address and phone number. How could this best be done? Please provide specific provisions for accomplishing these

Tillis also seems to recognize that one of the big problems of copyright law has long been the fact that it was designed in an era when the only real infringement and copying was done for commercial reasons, and that we've moved to an era when we all have super powerful copying machines in front of us (and in our pockets) all the time -- and that every single person with a computing device likely incidentally infringes thousands of times per day, based on a strict reading of the law. There's also the fact that copyright law treats a website run by a single user the same as a giant company like Google or Facebook. Tillis' suggested approach to dealing with this is to classify service providers by different markets or sizes:

The record established in my DMCA reform hearings indicated that an overarching principle of any reform should be making digital copyright less one-size-fits-all. The law needs to account for the fact that small copyright owners and small online services providers (OSPs) may have more in common with each other than they do with big copyright owners and big OSPs, respectively. Accordingly, I think we should consider whether copyright law should be revised to account for such differences among stakeholders. In particular, could copyright law borrow from employment law, or other relevant fields, to establish different thresholds for copyright owners and OSPs of different size, market share, or other relevant metric? If so, what is the best way to accomplish this? Is there a particular area of law, or existing section of the U.S. Code, that provides crucial guidance? As with all questions where it is relevant, please include in your response specific recommended legislative

Similar to the other questions mentioned, he's actually looking at the right issue, but approaching it with the wrong solution. This approach is almost certain to lead to weird incentives, potential regulatory capture to design specific rules, and a significant amount of gaming, as companies employ copyright lawyers to figure out how to stretch and twist around the rules' definitions to comply with the law as a technicality, but not in spirit. We've seen enough of that, and it leads to wacky scenarios like Aereo, in which a technically insane setup is put in place in the belief that it has to be built that way to comply with copyright law (even if the Supreme Court then stepped in with its weird "looks like a duck" precedent that made a mess of things).

Again, rather than trying to just craft a bunch of specific rules that will be gamed, shouldn't the focus be on setting an appropriate rules that actually makes sense across the board?

Where the Tillis questions get much more problematic is the suggestion that we should import some very dangerous, censorial and likely unconstitutional ideas like "notice-and-staydown" and full site blocking.

Section 512 places the burden on copyright owners to identify infringing materials and affirmatively ask the OSP to remove the material or disable access to it. This burden appears to strike the correct balance, but the burden that the notice-and-takedown system itself places on copyright owners is too heavy; the system is also woefully inefficient for both copyright owners and service providers. I believe U.S. copyright law should move towards some type of a notice-and-staydown system—in other words, once a copyright owner notifies a service provider that a use of a copyrighted work is infringing, the service provider must, without further prompting, remove subsequent infringing uses absent a statement from the user (whether the copyright owner or not) that they believe the use is licensed or otherwise authorized by law (e.g., fair use). What are your thoughts on such a system, and how could it best be implemented?

The injunctions available under section 512(j) have been narrowly interpreted by courts and thus little-used by copyright owners. Is it worthwhile for Congress to consider revising this provision to make injunctions more readily available for website-blocking in special circumstances (with an eye toward article 8(3) of the Information Society Directive)? Such injunctions could be issued by a special tribunal and appealed to federal district court, or, out of concern for user protections, the law could require that injunction orders come from the district court alone. If warranted, what would be the best way to enact limited website-blocking via such injunctions? Again, please provide suggested legislative text. If you do not think the law should be amended to expand the availability of injunctions, please be specific about any ways you think section 512(j) could be improved.

Yikes on both of these. Notice-and-staydown is a horrifically bad idea. We're already dealing with a situation in which the notice-and-takedown process is regularly and consistently filed for non-infringing reasons, with an effort to suppress or silence speech. Notice-and-staydown would make that problem much worse.

Furthermore, notice-and-staydown ignores the fact that it's not the content that is infringing, but the use. That is, the very same content may be infringing in some use cases, but not in others. The very same content could be fair use in certain contexts, but not in others. Notice-and-staydown ignores that.

Site blocking has massive 1st Amendment implications as well. Taking down an entire site because of some infringing material would violate the general principle as set by the Supreme Court that any regulation impacting speech must be narrowly tailored to only impact speech that is not protected by the 1st Amendment. Full site blocking does not do that.

There are other big reform ideas that seem to be missing entirely. A full (non-tinkering on the edges) reform bill should look at issues like false takedowns, copyright terms, digital 1st sale rights, and statutory damages. But those don't appear to be mentioned. There is one question that touches on the fact that a ton of takedown notices are bogus, but only loosely:

At the same time that Congress should revise section 512 to ensure that infringing material stays down once identified, it should also discourage the over-sending of notices as a counter-balance to the more significant action that an OSP must take after receiving a notice. This could be done, for example, by heightening the requirements for accuracy in notice sending, possibly with stricter requirements and heavier penalties. As noted above, the standard may be more lenient for small entities and individuals. How might the requirements be heightened in a meaningful way while not unduly burdening copyright owners trying to protect their work against infringement?

It seems odd that they don't even name the totally toothless 512(f) in this section, even though other questions focus on specific parts of the law. The idea that stronger penalties for false notices would "unduly burden copyright holders" is kind of laughable when you look at the current "balance" in copyright law, in which purely incidental infringement of a single thing could bankrupt most Americans.

In the end, while there are a bunch of bad ideas presented in the questions, I'm still somewhat pleasantly surprised that many of the questions raised indicate an awareness of the actual problems in current copyright law. That's better than we've seen in the past. That said, most of the proposed "solutions" will not help and may make things worse.

That is why it's important for people to weigh in with thoughtful ideas on how to actually fix copyright law. That said, there's an insanely short time frame for this. The letter asks for any comments to be sent in by December 1st (next week!) and says that Tillis intends to release a draft bill by December 18th. That seems like an astoundingly short time frame (1) for people to organize and submit their ideas and (2) for Tillis' staff to incorporate them into legislative text.

There's still a very strong chance that the bill that Tillis proposes will be a dangerous disaster towards free speech and innovation online. Certainly, some of the questions strongly hint at that. However, the very fact that the questions also indicate at least some legitimate recognition of the problems of today's copyright law is nice to see. Hopefully, Tillis and his staff recognize that the answers to these problems is not to make copyright law even more draconian, but to recognize that there are better solutions.

Filed Under: abusive takedowns, copyright, copyright reform, dmca, dmca 1201, dmca 512, dmca reform, free speech, innovation, notice and staydown, takedown notices, takedowns, thom tillis, triennial review

Ridiculous: 'Cyberpunk 2077' Will Ship With A Mode Just To Help Streamers Avoid DMCA Notices

from the thanks-twitch dept

You will likely have been following along with us as we have steadily commented on the ongoing controversy at Twitch. But if you're not read up on the topic, Twitch suddenly nuked zillions of hours of recorded content made by Twitch streamers in response to RIAA and game publisher DMCA notices, all without warning and all without a way to counternotice or get any of that content back. As the community went into revolt, Twitch continued taking down content, at times for sound effects within the games streamers were streaming. All the while, Twitch has issued a steady stream of apologies, while the streamer community has basically just shouted "Well then do something!" in response.

But Twitch hasn't done anything. Not a damned thing. Which means it's been left to the forward-thinking game publishers that actually realize how beneficial these streamers are to their own success to do something instead. To that end, it's both great that CD Projekt Red has announced the forthcoming blockbuster Cyberpunk 2077 will have a game mode dedicated to using stream-safe music for streamers... and completely ridiculous that the publisher even has to do something like this.

During the latest installment of long-form Cyberpunk commercial series “Night City Wire,” UK head of communications Hollie Bennett explained that the game will have a mode that will not only remove licensed music from the in-game rotation but replace it with music that won’t get creators’ channels zapped out of existence. Handy!

“If you’re planning on livestreaming Cyberpunk, or if you just want to make videos, we want to introduce you to a new mode that will allow you to disable certain copyrighted tracks,” Bennett said. “We know that for content creators, licensed music can sometimes be problematic. So with this new mode, you’ll be able to disable a small number of selected tracks which could cause some issues, replacing them with a different song—helping to avoid any problems.”

So because Amazon-owned Twitch couldn't be bothered to simply license the music in games in some sort of blanket manner, and because the music industry is so cartoonishly impermissive with its content, a game publisher has to step in to help. This comes as part of CD Projekt Red's long history of being fan and public friendly, but it really shouldn't have had to take such measures. Somebody somewhere along the way should have been on the side of the streamers who make these game products more popular, leading to more sales.

It’s still far from an ideal solution—nothing short of Amazon and Twitch striking a licensing deal with the music industry would be—but for now, it will have to do.

For now it will indeed have to do. But it sure would be nice if streaming platforms generally, and Twitch in particular, could get their collective heads out of their asses long enough to get their shit together and support their communities.

Filed Under: copyright, cyberpunk 2077, dmca, licensing, streaming, streaming safe mode, takedowns, video games
Companies: amazon, cd projekt red, riaa, twitch

Twitch's No Good, Very Bad Time Continues: Part 1

from the what-a-con dept

I'm beginning to wonder if the folks that run Twitch are secretly attempting to commit corporate suicide. The past several weeks have seen the popular streaming platform embroiled in controversy. It began when, in response to the RIAA labels DMCA attacks on streamers, Twitch took the unprecedented step to simply nuke a zillion hours of recorded content without warning its creators. In the wake of that, the platform kept essentially silent on its actions, simply advising its creators that they should "learn about copyright". In lieu of any real crisis communication, Twitch instead rolled out the release of a new emoji, pissing everyone off. Then came Twitch's apology, where the Amazon-owned platform acknowledged that it really should have had a method for letting streamers know which content was accused of infringement instead of nuking it all, while also continuing the DMCApocalypse, getting so granular as to allow streamers to be targeted by DMCA claims on game music and sound effects, including on videos that had already been taken down.

With its creators and patrons both in full revolt, it probably wasn't the best timing that Twitch's GlitchCon remote convention took place mid-November. Complaints about the convention were far-reaching, but much of it centered on the coin spent promoting it instead of Amazon simply licensing music so streamers could stream, along with the terse commentary on the turmoil itself.

We'll start with the promotion of the event.

The convention took place on November 14, but a difficult-to-ignore sensation of dissonance began to creep in before it even kicked off. To promote the event, Twitch sent themed trailers decked out with Twitch merch to select streamers—which streamers began tweeting about on November 13. While the streamers who’d received the vehicles seemed pleased, the response from many others was uniform: Why was Twitch spending money on glitzy trailers when it should’ve been putting every penny it could toward licensing music, thereby beheading the DMCA dragon currently terrorizing the platform?

Of course, the teams at Twitch that handle event planning and DMCA-related matters are very different, and this question ignores the reality of how budgeting tends to work at large companies. However, the broader sentiment from streamers was understandable; over the course of the past month, Twitch has massively eroded community trust by leaving streamers high and dry when the music industry finally came to collect its toll, forcing streamers to delete their entire histories instead of providing them with alternatives—or even accessible means of contesting copyright claims. During the lead-up to GlitchCon, streamers were not exactly in a celebratory mood.

As Kotaku notes, it's not entirely fair to simply claim that the money spent promoting GlitchCon should have been spent on music licensing instead. But it's not entirely unfair either, and the larger point is that Twitch did this to itself. By acting so callous with the work of its creators, and by then spending promotional budget dollars in a way that reminds everyone that this is a company backed by Amazon, it was inevitable that creators would throw up their hands in disgust. Whatever we might want to say about the imperfection of copyright laws, or the broken method by which copyright is policed at scale by platforms like Twitch, it most certainly is true that Amazon/Twitch could have avoided literally all of this by simply licensing a bunch of RIAA music. It's not like Amazon couldn't have afforded it. But, instead, Twitch's creators got screwed.

But when Twitch CEO Emmett Shear gave his keynote to kick off GlitchCon, the pushing of any information off to a future Q&A coupled with the highlighting just how bad a job his company did in supporting streamers felt like the worst of all worlds.

“It’s obvious that many of you want and deserve a lot more information from us, and a 10-minute Q&A session wouldn’t even come close to the level of depth of conversation that we want to have with you,” he said, noting that there will be a town hall devoted to the topic of DMCAs next month. He proceeded to apologize, largely reiterating what Twitch said in an apology letter it posted last week.

“If you receive a DMCA takedown, you should be able to know exactly what the content is or, if you believe you are authorized, you should know how to contest the takedown. I believe it’s a failing of our email to creators on October 20 that we didn’t include enough of this information, and it’s an issue with our current systems that we’re working to improve,” Shear said during the GlitchCon keynote. “We should have had better tools for you to manage your content, and we wish we did. We’re sorry those tools weren’t available when you needed them and that so many creators had to delete their videos capturing their communities’ best moments and accomplishments.”

Who this message was supposed to please is entirely unclear to me. Great, Twitch has acknowledged that it failed to support its creators with the tools necessary to do DMCA takedowns and restoration correctly. The first step to correcting a problem, as they say, is acknowledging you have a problem. But then announcing that the Twitch community deserves a ton of answers here, but they won't get them for another month? That's damned near self-immolation in the tech space. A glitzy convention put on without addressing a community in near revolt...why? Why in the world would you even take that virtual stage without being prepared to address the controversy?

It's not surprising that the reaction from the Twitch community was largely negative. And, because of Twitch's bullheaded approach to mostly ignoring all of this, that negativity overshadowed the rest of the convention, including some fairly positive happenings at Twitch.

Attempted corporate suicide is starting to look like a term bereft of exaggeration.

Filed Under: copyright, dmca, emmett shear, glitchcon, takedowns
Companies: twitch

GitHub, EFF Push Back Against RIAA, Reinstate Youtube-dl Repository

from the DEAR-RIAA-YOU-ARE-CORDIALLY-INVITED-TO-GFY dept

A few weeks ago, the RIAA hurled a DMCA takedown notice at an unlikely target: GitHub. The code site was ordered to take down its repositories of youtube-dl, software that allowed users to download local copies of video and audio hosted at YouTube and other sites.

The RIAA made some noise about copyright infringement (citing notes in the code pointing to Vevo videos uploaded by major labels) before getting down to business. This was a Section 1201 complaint -- one that claimed the software illegally circumvented copyright protection schemes applied to videos by YouTube.

The takedown notice demanded removal of the code, ignoring that fact there are plenty of non-infringing uses for a tool like this. It ignored Supreme Court precedent stating that tools with significant non-infringing uses cannot be considered de facto tools of infringement. It also ignored the reality of the internet: that targeting one code repository wouldn't erase anything from dozens of other sites hosting the same code or the fact that engaging in an overblown, unjustified takedown demand would only increase demand (and use) of the software.

Youtube-dl is a tool used by plenty of non-infringers. It isn't just for downloading Taylor Swift videos (to use one of the RIAA's examples). As Parker Higgins pointed out, plenty of journalists and accountability activists use the software to create local copies of videos so they can be examined in far more detail than YouTube's rudimentary tools allow.

John Bolger, a software developer and systems administrator who does freelance and data journalism, recounted the experience of reporting an award-winning investigation as the News Editor of the college paper the Hunter Envoy in 2012. In that story, the Envoy used video evidence to contradict official reports denying a police presence at an on-campus Occupy Wall Street protest.

“In order to reach my conclusions about the NYPD’s involvement... I had to watch this video hundreds of times—in slow motion, zoomed in, and looping over critical moments—in order to analyze the video I had to watch and manipulate it in ways that are just not possible” using the web interface. YouTube-dl is one effective method for downloading the video at the maximum possible resolution.

At the time, GitHub remained silent on the issue, suggesting it was beyond its control. Developers who'd worked on the youtube-dl project reported being hit with legal threats of their own from the RIAA.

There's finally some good news to report. The EFF has taken up GitHub/youtube-dl's case and is pushing back. A letter [PDF] from the EFF to GitHub's DMCA agent gets into the tech weeds to contradict the RIAA's baseless "circumvention" claims and the haphazard copyright infringement claims it threw in to muddy the waters.

First, youtube-dl does not infringe or encourage the infringement of any copyrighted works, and its references to copyrighted songs in its unit tests are a fair use. Nevertheless, youtube-dl’s maintainers are replacing these references.

Second, youtube-dl does not violate Section 1201 of the DMCA because it does not “circumvent” any technical protection measures on YouTube videos. Similarly, the “signature” or “rolling cipher” mechanism employed by YouTube does not prevent copying of videos.

There's far more in the letter, but this explains it pretty succinctly in layman's terms:

youtube-dl works the same way as a browser when it encounters the signature mechanism: it reads and interprets the JavaScript program sent by YouTube, derives the “signature” value, and sends that value back to YouTube to initiate the video stream. youtube-dl contains no password, key, or other secret knowledge that is required to access YouTube videos. It simply uses the same mechanism that YouTube presents to each and every user who views a video.

We presume that this “signature” code is what RIAA refers to as a “rolling cipher,” although YouTube’s JavaScript code does not contain this phrase. Regardless of what this mechanism is called, youtube-dl does not “circumvent” it as that term is defined in Section 1201(a) of the Digital Millennium Copyright Act, because YouTube provides the means of accessing these video streams to anyone who requests them. As a federal appeals court recently ruled, one does not “circumvent” an access control by using a publicly available password. Digital Drilling Data Systems, L.L.C. v. Petrolink Services, 965 F.3d 365, 372 (5th Cir. 2020). Circumvention is limited to actions that “descramble, decrypt, avoid, bypass, remove, deactivate or impair a technological measure,” without the authority of the copyright owner… Because youtube-dl simply uses the “signature” code provided by YouTube in the same manner as any browser, rather than bypassing or avoiding it, it does not circumvent, and any alleged lack of authorization from YouTube or the RIAA is irrelevant.

GitHub's post on the subject explains the situation more fully, breaking down what the site's obligations are under the DMCA and what it does to protect users from abuse of this law. It also states that its overhauling its response process to Section 1201 circumvention claims to provide even more protection for coders using the site. Going forward, takedown notices will be forwarded to GitHub's legal team and if there's any question about its legitimacy, GitHub will err on the side of USERS and leave the targeted repositories up until more facts are in. This puts it at odds with almost every major platform hosting third-party content which almost always err on the side of the complainant.

And the cherry on top is the establishment of a $1 million legal defense fund for developers by GitHub. This will assist developers in fighting back against bogus claims and give them access to legal advice and possible representation from the EFF and the Software Freedom Law Center.

Youtube-dl is back up. And the RIAA is now the one having to play defense. It will have to do better than its slapdash, precedent-ignoring, deliberately-confusing takedown notice to kill a tool that can be used as much for good as for infringement,

Filed Under: circumvention, copyright, copyright 1201, copyright 512, counternotice, recording software, takedowns, youtube-dl
Companies: eff, github, riaa, youtube

Twitch Continues To Trip Over Itself In Response To DMCA Apocalypse

from the do-better dept

What a few weeks for Twitch. You will recall that the platform went about pissing a ton of its talent and viewers off by nuking a metric ton of video content on the site in response to a flood of DMCA takedown notices, most of them from the RIAA. And this truly was the nuclear option, far different from the notice/counternotice system most platforms use. In fact, it was so extraordinary that it arguably lost Twitch its DMCA safe harbor. Regardless, when the company then followed up with a message to all Twitch creators that they should go educate themselves on matters of copyright and proactively delete any recordings or clips that might run afoul of copyright law, it created a cluster-fuck with virtually nobody having any idea how or what they should be doing. In response to the turmoil, Twitch brilliantly rolled out an announcement for a new emoji.

And it just keeps getting worse. This week, Twitch has finally come out with an apology to its talent, noting that the company, bought by Amazon in 2014, probably should have been able to provide better tools and a system that wouldn't have required the mass deletion of millions of hours of recorded content.

“You’re rightly upset that the only option we provided was a mass deletion tool for Clips, and that we only gave you three-days notice to use this tool,” the company wrote. “We could have developed more sophisticated, user-friendly tools awhile ago. That we didn’t is on us. And we could have provided creators with a longer time period to address their VOD and Clip libraries—that was a miss as well. We’re truly sorry for these mistakes, and we’ll do better.”

Twitch also committed to developing additional tools that will hopefully grant streamers more granular control over their recorded content, audio, and reviewing/contesting copyright claims, but it did not provide a release date for those much-needed features. It went on to try and explain why it doesn’t just obtain music licensing rights like Facebook has for its livestreaming platform, but while it said that those solutions won’t work for Twitch in particular, it stopped short of explaining exactly why.

So, apology with a side of obfuscation that doesn't actually put creators in any better a place than they were yesterday. There's still no way to counternotice. No commitment from Twitch to supporting creators' rights when it comes to fair use. And the explanation that a licensing deal with the music labels would take too much money away from Twitch creators considering how sparingly they use music in their streams comes along with two problems. The first problem is that the statement leads to an obvious question: You're Amazon; why don't you just eat the licensing costs and let us create?

But the second problem is that the DCMA apocalypse is continuing and it's starting to get ridiculously granular.

This comes during an especially turbulent week on the DMCA front. In the past few days, streamers have reported getting targeted by copyright claims and Twitch’s automated systems for music and sound effects in games, as well as clips they’ve already deleted. One streamer, MichalRonin, had his audio auto-muted for broadcasting a wind gust sound in World of Warcraft.

“Only music I’ve had on stream was in-game WoW music, played by the game itself,” MichalRonin wrote on Twitter yesterday. “Yet I’ve got ‘muted audio’ on the latest VOD, apparently in Sen’Jin village.” He then posted a screenshot that mentioned a “Medium Wind Storm with Gusts, Whistles and Low Rumble” sound effect from the Hollywood Edge Sound Effects Library.

This example was one of many. If DMCA notices are suddenly going to start coming in and being acted upon over video game sound effects, never mind the game music that comes along with streaming a let's-play, then Twitch is essentially over. On top of that, some streamers report that even covers of video game music are getting caught up in the DMCA takedowns. Still others received strikes for content that had already been deleted.

Streamer JasonParadise deleted all of his clips on October 23, the day Twitch resumed regular DMCA processing after holding back thousands for a handful of months.

“What the fuck was the point of deleting all of my VODs/clips back on October 23rd? The strike on an old clip (that no longer existed) came in ten days later,” he wrote on Twitter.

Questions to Twitch about all of the above by the media have gone entirely unanswered as of the time of this writing. It all paints a picture of a popular and well-traveled platform with a parent company that ought to be able to provide all of the capital, platform tools, and public messaging in the world... instead having none of that.

And with YouTube once again making a strong play for these kinds of streamers, one wonders just how long it all can last.

Filed Under: clips, dmca, music, takedowns
Companies: riaa, twitch

Netflix Gets Cute Using DMCA Notices To Take Down Tweets Critical Of 'Cuties'

from the too-cute-by-half dept

Cuties, the stupid non-controversy against Netflix that simply will not go away. The film, which won awards at international film festivals, centers on a pre-teen and is a coming of age story about a young lady growing up in both a strictly conservative upbringing combined with living in the hyper-sexualized Western culture. While the whole story is about this juxtaposition, Netflix rather stupidly promoted the film using images that focused on the latter. The result was chaos, with large swaths of Puritan-Twitter screaming about boycotting Netflix entirely and one pandering prosecutor in Texas bringing an indictment against Netflix for promotion of lewd visual material depicting a child.

With everyone very quickly lighting themselves on fire over an award winning film, Netflix sat back and calmly explained what the film was about and why it had... just kidding, Netflix is now out here issuing DMCA takedowns for those tweeting critiques of its decision to distribute the film.

Netflix's takedown requests, which are still rolling in today, seem only to have targeted tweets that described the film negatively, although some more than others.

"IMAGINE A CHILD SEEING THIS #Cuties #Netflix #CancelNetflixCuties," one message read. "WARNING CONTAINS EXPLICIT CONTENT," another similar message said. "Go ahead and try to justify how this film is an appropriate representation of 11 year olds. I'll wait. #CancelNetfilx."

Some of the dozens of tweets Netflix issued DMCA claims against used clips from the actual movie, TorrentFreak reports, in which case Netflix's claims are understandable. However, many of the tweets in question shared the film's trailer, which is widely and publicly available on YouTube for anyone to view or share.

DMCA takedowns of trailers, as we've explained before, never make sense. Ever. Ever ever. What Netflix is doing instead is lay bare its intentions behind these takedown notices, which are obviously centered on attempting to censor critical commentary around its decisions surrounding the film. This becomes especially apparent when put in the context for how and for what Netflix has, in the past, bothered issuing DMCA takedowns.

TorrentFreak notes that the cluster of claims is unusual for Netflix, which has sent roughly 300 DMCA claims to Twitter in the past month, half of which centered on tweets related to Cuties. Before Netflix started targeting Cuties tweets, most of the claims it sent were related to accounts known for distributing pirated content.

And so the Streisand Effect kicks in. By trying to bury criticism, the public becomes all the more aware of that criticism. By trying to censor a controversy that was probably juuuuuuust about to go away, instead it gets recycled back into the news cycle.

Netflix, tech company as it is, should absolutely know better. Reliant on the First Amendment as it is, it should absolutely not be taking actions like this that tamp down speech. And given that Netflix is not entirely without blame for the controversy in the first place, it sure would be nice if the company demonstrated skin thick enough to take a little heat now and again.

Filed Under: criticism, cuties, dmca, overreaction, streisand effect, takedowns, trailer
Companies: netflix

Twitch's Freak Out Response To RIAA Takedown Demands Raises Even More DMCA Questions

from the not-how-any-of-this-is-supposed-to-work dept

As many of you probably saw last week, Twitch decided to delete a ton of videos in response to DMCA takedown claims (which most people believe came from the RIAA). As we pointed out earlier this year, the RIAA had started flooding Twitch with DMCA takedowns over background music used in various streams. The whole thing seemed kind of silly, and now it appears that Twitch (despite being owned by Amazon and having some pretty good lawyers) was caught without a plan.

And that manifested itself in the way it handled these takedowns. Rather than the standard process -- taking the content down, letting the user counternotice, and then potentially putting it back up 10 days later if no lawsuit was filed -- Twitch decided to just totally wipe those files out and not even leave an option open to users to counternotice.

The key bit:

We are writing to inform you that your channel was subject to one or more of these DMCA takedown notifications, and that the content identified has been deleted. We recognize that by deleting this content, we are not giving you the option to file a counter-notification or seek a retraction from the rights holder. In consider of this, we have processed these notifications and are issuing you a one-time warning to give you the chance to learn about copyright law and the tools available to manage the content on your channel.

We know that copyright law and the DMCA are confusing. Over the past few months, we've been improving the tools available to help you manage music use in your live and recorded content. These include the ability to delete all of your Clips at once and control who can create Clips on your channel, scanning new Clips with Audible Magic and launching a free way to stream high quality music on your channel, Soundtrack by Twitch. Now that these tools have been released to all creators, we will resume the normal processing of DMCA takedown notifications received after 12 noon PST on Friday, October 23, 2020.

So... reading between the lines here, it sounds like someone (likely the RIAA or some similar organization) ratcheted up the threats for not being responsive enough on takedowns -- and someone up top just said "nuke 'em all, so we can claim we got rid of everything." But that's incredibly stupid on multiple levels.

That first paragraph above is completely nonsensical and I've read it over multiple times trying to parse out what the hell it means. The company admits that it isn't giving anyone any chance to counternotice, or get takedown demands removed. In fact, it sounds like they can't even bring back any of this content. It's just gone. That could really suck for some users who will not have that content at all.

But then in follows up with "In consideration of this..." which sounds like it's apologizing and going to give users back some sort of benefit... but instead, it says it's giving those users "a warning" and telling them to "learn about copyright law." What? What sort of absolute nonsense is that. If someone's video was mis-identified, or the video was fair use, why should they get a warning and have to "learn about copyright law"? It feels a lot more like Twitch should learn about copyright law.

Indeed, there's a potential argument that by deleting the videos and not allowing there to be a counternotice, Twitch got rid of the DMCA's safe harbor. If you read Section 512(g) of the DMCA, about the replacement of removed or disabled material, it says that there is no liability for the removal of content (which makes sense) with three "exceptions" delineated in Section 2. Here's the relevant part of 512(g):

(1)No liability for taking down generally.—
Subject to paragraph (2), a service provider shall not be liable to any person for any claim based on the service provider’s good faith disabling of access to, or removal of, material or activity claimed to be infringing or based on facts or circumstances from which infringing activity is apparent, regardless of whether the material or activity is ultimately determined to be infringing.

(2)Exception.—Paragraph (1) shall not apply with respect to material residing at the direction of a subscriber of the service provider on a system or network controlled or operated by or for the service provider that is removed, or to which access is disabled by the service provider, pursuant to a notice provided under subsection (c)(1)(C), unless the service provider—

(A)takes reasonable steps promptly to notify the subscriber that it has removed or disabled access to the material;
(B)upon receipt of a counter notification described in paragraph (3), promptly provides the person who provided the notification under subsection (c)(1)(C) with a copy of the counter notification, and informs that person that it will replace the removed material or cease disabling access to it in 10 business days; and
(C)replaces the removed material and ceases disabling access to it not less than 10, nor more than 14, business days following receipt of the counter notice, unless its designated agent first receives notice from the person who submitted the notification under subsection (c)(1)(C) that such person has filed an action seeking a court order to restrain the subscriber from engaging in infringing activity relating to the material on the service provider’s system or network.

So it certainly appears that you can read this to say that Paragraph (1) (the no liability bit) "shall not apply... unless the service provider... replaces the removed material and ceases disabling access to it not less than 10, nor more than 14, business days following receipt of the counter notice...." And thus, there is a reading of this that says that not replacing the material after receiving a counternotice (that is not followed by an actual lawsuit) could remove the safe harbor protections.

Of course, this may not really matter for a variety of reasons. First off, who would actually sue here? The safe harbors are from the copyright holder, but here the copyright holder is likely to be happy that Twitch has gone overboard in deleting all of the content. Second, this aspect of the safe harbor, which can be read to require the replacement of content following a counternotice has some other problems in that it can be read as forcing a company to host content. And, a website should have the freedom to not host any content it doesn't wish to host, even if that content is not infringing.

I vaguely recall a lawsuit a few years ago on this point, though in search for it I can no longer find it. From my (apparently faulty) memory, I recall that someone challenged a website's unwillingness to restore content after a counternotice, and the court found that a website has every right to keep a work it disabled offline. If anyone can remember which case this is, let me know.

Still, it remains somewhat perplexing that this is how Twitch handled all of this, and then claiming it would go back to "normal" DMCA processing as of the end of last week. Why couldn't it have just kept that up this whole time?

Filed Under: counternotice, dmca, takedowns
Companies: riaa, twitch

While Trump Continues To Complain About 230, It's Copyright Law That Once Again Actually Gets His Content Removed

from the why-is-no-one-looking-at-fixing-copyright? dept

Once again this week, the President decided to attack Section 230 because social media companies decided to highlight that he was posting dangerous misinformation (this time about the relative dangers of COVID-19, which he was downplaying). Yet, for reasons I do not understand, the President never seems to address copyright law, even though that law is what is actually forcing his and his campaign's content to be legally removed from social media.

Over the last few months we've highlighted multiple times that Trump and his campaign have had posts removed from social media due to DMCA 512 takedown notices. And it happened again this week after Twitter removed a tweet from the Trump campaign on copyright grounds.

The tweet in question made use of a video clip showing the San Francisco 49ers wide receiver Brandon Aiyuk scoring a touchdown from last Sunday's 49ers/Eagles game. The clip (somewhat ridiculously) superimposed Trump's head over Aiyuks, and put the well known graphic of the coronavirus on Eagle's defender Marcus Epps (whom Aiyuk leapt over in getting to the end zone). The message of the video (stupidly) is that Trump was somehow able to "avoid" the coronavirus (which, I should remind you, he did not). The video is stupid on multiple levels, including the the sickening and despicable implication that those who died from COVID-19 are somehow just not strong enough.

However, I think there's a pretty strong argument that the video would be protected as fair use -- and that the takedown issued by the copyright holder (likely the NFL) was not a legitimate takedown. In fact, it's possible that the NFL issued the takedown for political reasons, as there's no argument that this somehow harmed the NFL directly. It's a short clip. It's used in a transformative (if stupid) way.

In other words, this is an actual example of the law being used for censorship. Unlike Section 230. And yet, we don't see Trump or his supporters calling for that aspect of copyright law to be fixed. Indeed, copyright law is even worse, because if the Trump campaign keeps getting copyright strikes like this, the law says that Twitter must shut down his account for repeat infringements. Will Trump and his supporters finally see that the real problem for censorship is copyright?

Instead they're asking to take away Section 230, which (at best) would create a situation more like copyright in which the legal incentive is much stronger towards pulling down such content. It remains incredible to me that in all of these discussions about social media and "censorship" everyone is focused on the law that protects speech online, rather than the law that forces websites to pull down legal content.

Filed Under: censorship, content removals, copyright, dmca 512, donald trump, fair use, parody, repeat infringer, section 230, takedowns, trump campaign
Companies: twitter