Copyright Office 512 Roundtable: Notice and Takedown

A couple of introductory thoughts: There are over 20 panelists per hour and a half panel.  The math on that is not good in terms of substantive contributions, even if no one rolls out their talking points at all in response to specific questions (not bloody likely, and I don’t exclude myself, but maybe the Office has figured out an issue that has bedeviled moderators since the beginning of conferences; if it involved a bucket of water, that could even be fun to watch).  After participating in the PTO/NTIA green paper process and the associated best practices in takedown receipt process, I am fairly confident that there’s not much room to move among the different interests.  I take it that we are creating a record for a potential Copyright Office proposal.  If so, the initial topics offered are not promising ones.

 

Copyright Office: Jacqueline Charlesworth, Karen Temple Claggett, Kim Isbell, Brad Greenberg, Cindy Abramson, Rachel Fertig

 

SESSION 1: Notice-and-Takedown Process—Identification of Infringing Material and Notice Submission

 

Official description: Overall effectiveness of process, including issues relating to timing and linked content; volume of notices; burdens on large- and small-scale creators and copyright owners; considerations and strategies in sending notices, including automated notices versus human review; fair use considerations; moral rights issues; “whack-a-mole” issue; relevant empirical data; and other pertinent issues.

 

Charlesworth: 512 is approaching the 20-year-mark, and not merely of interest to © nerds or particular industries. Affects anyone who interacts w/the internet.  Dramatic impact on the way the internet and content services evolved in the US.  Impact is growing—astonishing that Google is on track to receive 1 billion notices of alleged infringement this year.  Exponential number may not have been apparent in 1998 when Congress was looking at bulletin boards. What does that mean?  Good, bad?  Struck by comments: wide chasm b/t those who perceive system as essentially working beneficially and those who see it as seriously flawed.  Dickens: Tale of Two Cities—It was the best of times, it was the worst of times. Age of wisdom, foolishness; spring of hope, winter of despair.  Hope: exchange of views will lead to more hope than despair. Need to find common ground. 

 

Note: I can’t read most of the nametags and will not necessarily be able to ID all speakers.

 

JC: Those who send notices, how is the system working for you?

 

Lisa Hammer Independent Film Director: 3 takedown notices for a film just sent to film festivals; takes up time that could be devoted to film.

 

JC: how do you do that?

 

Hammer: YT I get 3-4/week; contact them and takedown. They’re good about it but timeconsuming. Other platforms, torrents.

 

JC: How do you search?

 

Hammer: My partner gets in contact w/ me and asks for takedown. She’s looking while I get editing done. She brings it to my att’n.  Someone told me to type in name of my film next to torrent and I found hundreds, though it’s not even released.

 

JC: How did copies get out?

 

Hammer: film festival website uploads may have leaked them.

 

Q: how much time?

 

Hammer: Daily.

 

Natalie Madaj National Music Publishers’ Association: Timeconsuming and ineffective.  Send on behalf of publishers large and small. Many less than $50,000/annual revenue. Don’t have resources to send notices on their own.  Looked into automated processes, but six-figure licensing fees are quoted. And Lenz would require manual review, which probably requires an att’y.  Not ok for other publishers who’d need to hire an att’y.

 

JC: NMPA offers this to its constituents?  How much staff time/resource level?

 

Madaj: small; I’m the only att’y + an intern who spends her time searching the internet. We look for sheet music, arrangements; also on behalf of other publishers, we look at streaming services.

 

Claggett: what’s your notice volume?

 

Madaj: for one sheet music site, ustore.com, 13,000 URLs, and they still haven’t shut down or implemented a repeat infringer policy but we don’t have resources to file a lawsuit. [And you would if 512 didn’t exist?]

 

Ellen Schrantz, Internet Ass’n: Robust success: the most fundamental point is that w/o that law there’d be no expeditious removal; you’d still have the task of removing content but you’d have to sue to get it down w/o 512.  Volume of notices is indicator of success. 

 

JC: safe harbors; w/o 512 the ecosystem would be different.

 

Schrantz: Yes, the organization of the internet would be different—shared responsibility. When you talk about removal of infringing content; w/o 512 the system would be work.  High incentives for provider participation exist now and should be protected

 

Maria Schneider, Musician: I face all different kinds of infringement: recordings, performances of my work, print music, instructional videos I sell on my site.  Whack a mole.  Showing up on torrent sites where I have no DMCA access.  Foreign sites that there’s nothing I can do about it.  Sites that in order to check for takedowns, I have to check the download and face the possibility of getting infected by a virus. Odds stacked against me. I’d do Content ID on YT but they don’t accept me b/c I’m not big enough and don’t want to give my catalog to monetize. Game is stacked for us to lose.  I would like to do other things. I don’t have time/$ to do this; most of us just give up and go make something else. Sick game.

 

George Johnson, Geo Music Group: It’s absolutely not working. It’s only working for Google and licensors. Not working for creator. We don’t respect the exclusive right of ©.  Constitution/§106; it’s the © Office’s job to protect our exclusive right.  DMCA implemented WIPO rules on ©, but not my exclusive right or constitutional supremacy.  European Union/UN rules destroy American ©. Talk all you want about safe harbors/notice and takedown; we should abolish 512 and start putting ISP people in jail. That’s our problem, not notice and takedown.

 

Eugene Mopsik, Am. Photo. Artists: we’re the poster children for whack a mole. Reappearance of images after notice and takedown is startling.  Friend who does it regularly said, until recently, when Google coded their field for URL, they wouldn’t allow you to paste; had to manually type it into the space until recently, though it has been changed, to Google’s chagrin he’s sure.  Amount of resources photogs have to dedicate to adequately police: would be fulltime job. Most of our members are 1-2 person studios w/o that manpower; mostly they can’t even ID who they’re after. Beyond that, the Lenz decision has had a chilling effect on notice and takedown b/c photogs couldn’t evaluate fair use.  Even when you do ID the infringing uses, there’s no adequate resource for visual artists b/c there’s no small claims.

 

Claggett: CO has looked at visual artists—is there something unique about visual art in terms of §512.

 

Mopsik: It’s the sheer volume. Digital artists create more works on a daily basis than any other genre. Ease of abuse. Too easy to retransmit and display images w/o attribution, to strip metadata; there’s no ease of enforcement.

 

Alisa Coleman: ABKCO Music & Records: 512 is broken. Too much time figuring out ways to find where our content is being used.  Alternate methods—we’re a small company w/ valuable content and we have to download apps to see if people are using our content. Have to search through for correlation b/t compositions and master recordings.  We have 2 people working almost constantly to police/detect.

 

JC: does the effort yield sufficient returns?

 

Coleman: Not necessarily b/c it pops up somewhere else. It’s a constant battle.

 

Richard Burgess, Am. Ass’n Independent Music: Pretty tough 15 years for music. Many services are now paying reasonably well, and streaming can make the industry recover.  B/c of people hiding behind 512, we can’t control our content to the extent we need to in order to be a good business again. We represent 100s of independent labels and more artists; most have just given up sending takedown notices.  At the lower end, simply don’t have the resources.  Even if you do, it’s whack a mole.  Particularly shocking when abuses are coming from company w/tech resources to solve this problem, as they do w/porn.  [They do? Are you sure you have SafeSearch turned off?]  We need notice and staydown, not takedown.

 

Deborah Robinson, Viacom: At Viacom, we thrive on fan engagement, but still spend lots of time, resources, $: we use multiple people, vendors, proactive measures, reactive measures—both manual and automated processes.  [? Couldn’t understand] We focus on full content; other means, numbers might be different. Room for collaboration on tech that would help with filtering or staydown.

 

JC: give us more of flavor of how much of your takedown effort is automated v. human and what interaction is?

 

Robinson: even though we have automated processes, a lot require second level manual review. We spend a great deal of time manually looking for content once the automation has identified candidates.

 

JC: do you do a first cut through automated process? 

 

Robinson: yes.  Oftentimes.

 

JC: resources to entire process?

 

Robinson: great deal of money and time: in house and vendors.

 

Claggett: focus more on full length content—is that time/resource question for prioritizing, or is that something else?

 

Robinson: b/c we want to be fair. We want fans first; provides better tolerance for fair use. Plus it takes lots of time to look at content other than full copies.

 

JC: You mean fair use review.

 

Jonathan Band, Amazon: Amazon is on multiple sides—creates and distributes award-winning content; provides platforms for others to distribute content; hosts web content, AWS leading cloud service providers; receives and sends takedowns.  Amazon feels system is working; reasonable compromise—yes, it’s burdensome to ID and find infringing content.  OTOH, it’s burdensome to respond to takedowns. But it’s better than any possible alternative. So overall the view is that it works.  Also, focusing on small creators: Amazon’s system works efficiently and effectively for small creators who feel their content is being infringed; responds expeditiously. The balance provided by the DMCA allows Amazon to offer services/infrastructure to allow small startups to provide streaming services.  New distribution models can exist thanks to 512—w/o the DMCA, those services might not exist or Amazon might not be able to allow them to use Amazon’s infrastructure. New business models have emerged and DMCA gives a framework that enables them. 

 

Not on behalf of Amazon, but as lawyer who is representative of university press: their publications are sometimes online w/out authorization. DMCA is amazing: client says this book is appearing; I send the notice, and the next day it’s gone.  Client thinks I’m brilliant; it’s easy for me to do.  In that context, it works incredibly well.

 

JC: What resources at Amazon go to locating infringing content? What content is Amazon worried about and how is it identified? Is it automated?

 

Band: Amazon Prime distributes original video content, among other kinds of content; also has its own publishing arm.  That’s what it looks for, especially the video.  Resources: will have to get back to you. 

 

Claggett: For academic clients, process was easy.  Do you need to teach clients the law? Is there anything different about academic clients?

 

Band: good Q. © system touches many kinds of works and creators.  Certainly academic works aren’t in as much demand as Taylor Swift; so you don’t necessarily have as many sites that would be infringing, though the number of works total might be larger.  Easier to target.  But they’re © owners like anyone else.

 

David Kaplan, Warner Bros. Entertainment: We too focus on full length content; overwhelmingly our enforcement is that—TV, interactive games, films. Only exception is pre-release content.  [I note that hasn’t been our experience, but that may well have been a contractor problem.]  We’ve seen increase in efficiency over past 5 years—faster response times, easier to submit; not equal emphasis on efficacy. Millions of notices isn’t measure of system working; could be measure of system not working. Whack a mole.  Take the point that it works super well in some systems. Genie can be put back in the bottle b/c no one was really interested in the work in the first place. That’s happened for us, like prerelease content that we’ve been able to stop proliferating—1 or 2 examples in 20 years. Popular content though spreads like wildfire. Once it’s out, that’s it. Staydown would be key. We sent 25 million notices last year. Only about 1-2% notices were search related. Vast majority weren’t.

 

Automated tools: there’s human review or human processes in the setup of the automation. Differs v. streaming, P2P, hosted—some tech/automation as well as human review. Depends on what kind—w/r/t search, we do this combining in house and 6-10 outside vendors specializing in niche parts of online piracy.

 

JC: Is it worthwhile to invest company resources to do this?

 

Kaplan: to a certain extent. We try to create clean, legit experience for online consumption. One way is to try to make sure that if people click on links, the content on pirate sites isn’t there. Injecting friction into the system works to our benefit. We’ve had some success, but there’s a lot more that could be done: staydown. We’ve seen pirate businesses be adept at stacking URLs so the user experience is continuously positive b/c the hosts invariably have some version of the film available almost continuously.  [Wouldn’t those sites already be outside the safe harbor?]

 

JC: are you able to see into the next one down the stack? Your notice addresses the top of the stack.

 

Kaplan: can’t see it at the same time, not until the first is taken down.

 

JC: is it automatically replaced?

 

Kaplan: we think so, b/c it’s so quick. 

 

JC: how prevalent is that, in your experience?

 

Kaplan: increasingly prevalent.

 

Claggett: do you send out another notice?

 

Kaplan: Yes.

 

Isbell: you mentioned 1-2% are sent to search. What types of services receive most?

 

Kaplan: close to 17 million were to ISPs with P2P infringements.  5 million were to hosting sites where you could download. 3 million were to streaming sites.

 

JC: has stacking ever come up with large online OSPs? What can you do more efficiently?

 

Kaplan: unfortunately, it’s largely a black box on the other side. We’re often told they’ll look at the issue and see if they can make more difficult for people to have 100s of different accounts, but despite talking about this for a year or two we haven’t seen progress.

 

Claggett: are there incentives or disincentives in terms of 512 for that collaboration?

 

Kaplan: disincentive for platforms to take effective actions to address it.

 

Claggett: why?

 

Kaplan: they don’t want liability.

 

Kathy Garmezy, Directors Guild of America: all our members are concerned; some have resources of studios, but not independent directors who often own their own ©.  Find out film has been pirated when someone they know tells them. Don’t usually have lawyers.  Problem of reappearing right away is very discouraging. Want staydown.  Everybody thinks of the sites as being the larger sites; sometimes our members are threatened because of notices.  Not every site is a big aggregator that will not threaten you even if it resists takedowns.

 

JC: forward contact information, sometimes personal, for individual creator—is that how the threats are being communicated?

 

Garmezy: anecdotal—it’s within the takedown process. 

 

JC: so the poster uses the personal info to issue the threat.

 

Garmezy: say that we know where you are, I won’t take it down, don’t bother me again.

 

Janice Pilch, Rutgers U Libraries: Disclaimer: these are my own opinions, not those of Rutgers or any library.  Even in universities, it’s quite obvious that there’s a 512 problem. One of main problems is user-generated “course learning platforms” that are making money from infringed academic content. Affects instructors who use their own © materials and then find them on the open internet; find them by accident. Uploaded by students to user-generated sites that aggregate and monetize course material. Affects ability to reuse their own course material, requiring reinvention of courses – exams, etc.  Affects their credibility when associated with piracy. Many instructors don’t understand how this happens. Takes time and effort to file; may not file notices on 3d party works; representative lists aren’t accepted.  Sites undermine course material and exams.  [GU wisely doesn’t let me reuse an exam anyway because students create exam banks and did so long before the internet.]  Takes a lot of time and anxiety and often the takedown doesn’t work.  Commercial so-called scholarly networking tools that aggregate scholarly works.  Pirate sites like SciHub, over 147 million scholarly articles.  [Already pretty clearly ignoring 512 and US law. What would change with staydown?]

 

JC: reference to HEOA: talk about that? Standards about educating students about copyright and plans for addressing online infringement. Is that helpful?

 

Pilch: That requires universities to establish © policies, and they’ve generally done that.  But extent to which people read the website—that’s a different story. My emphasis in my comments wasn’t on infringement by faculty members; often I think students don’t realize that what they’re doing is wrong. We do need more education.

 

Claggett: What’s the scope of the problem w/academic content? Do universities assist their professors with sending notices?

 

Pilch: only Rutgers—I don’t hear about all these situations. Qs might go to university counsel, some other administrator. Hard to say the scope. When it happens, tends to be upsetting, burdensome, timeconsuming. People are surprised that aggregation and monetization is allowed.

 

Claggett: can they ask the university for help?

 

Pilch: University counsel, IT dep’t, or me, but they have to construct notice themselves.

 

Victoria Scheckler, RIAA: Google’s receipt of billion notices is failure of DMCA. We’ve sent 105 million notices and continue to see members’ works show up again and again on these sites. 2014—278,000 notices to a site that claims DMCA safe harbor; 94% were for previously noticed work.  We are running into the stacked URL problem—multiple URLs go to the same content, and they only take down one URL, not all of them; they say one might have authorization but we didn’t authorize the site.

 

JC: Full length work might be authorized for someone and not someone else: how do you manage the who’s licensed issue?

 

Scheckler: full length works. We review sites on a daily basis.  Look for full scale infringing sites which shouldn’t benefit from DMCA but claim nonetheless. We check with our labels.

 

Lisa Shaftel, Graphic Artists Guild: individual creators—see Schneider and Mopsik—the most infringed works are music and images; some sites exist solely to infringe images, such as Pinterest. Everyone in this room has infringed cartoons they thought were funny by passing them around. Metadata is stripped by many sites, so even when visual creators make effort to ID themselves they can’t—especially on social media. Infringers crop image to remove name, or they include attribution presuming that attribution makes their use ok. Creators can’t satisfy ISP requirements for takedowns: many require that they prove © registration as part of takedown, or other means of proving ownership of image. Majority don’t and won’t register their works.  We create more images daily than any other creators; many can’t satisfy the ISP’s requirements. Even those w/registrations: the certificate doesn’t have a thumbnail of the image.

 

JC: But complying w/DMCA just requires swearing to authorization from © owner.

 

Shaftel: Sites make up their own requirements.

 

JC: What about pushing back and saying that the notice complies?

 

Shaftel: Who’s policing what individual thieves or webhosts do? FB’s takedown requirements are really convoluted. Every ISP is different. Most artists give up.  Some have required registration to prove ownership of the image, which isn’t in the statute.

 

Claggett: adding addit’l requirements would take them out of the safe harbor, presumably.  Are the artists suing?

 

Shaftel: No, it’s not possible. If the work isn’t registered, we have no recourse—ties into need for © small claims. [But not to need for change in 512.]  © protection lasts only until the work is posted on the internet. In reality, impossible for illustrator to grant exclusive license to image online. 

 

Claggett: in other contexts, people have talked about DMCA plus mechanisms like Content ID.  Have you seen any type of addit’l measures or licensing or DMCA-plus activity for image sites?

 

Shaftel: yes and no. Plus licensing system exists.  A lot of websites whose entire business model is to get viewers to put images on site; they sell ads or images.  They don’t want to comply; this is their business model.

 

Sandra Aistars Arts and Entertainment Advocacy Clinic, George Mason University School of Law: speaking for independent artists; agree w/others. Artists who we work w/ are making daily decisions about whether to create or to enforce © interests. Don’t have the same tools as larger corporate owners. Not afforded access to ©-plus or DMCA-plus tools like Content ID. Varied requirements from websites; addit’l hurdles to submit notice.  Not much improved.  One issue: b/c of imbalance in info disclosed about artist when she submits takedown v. what’s disclosed about the user who’s posted, artists are frequently afraid to send DMCA notices for fear of retaliation. W/r/t noted photojournalist who works in conflict areas: taken captive by Somali warlords, so not meek; her images are often taken and misrepresented. She’s taken a Kosovo image and misrepresented by revolutionary group as misrepresenting something else. She’s fearful of disclosing personal info to a group that potentially employs radical tactics. 

 

Greenberg: Concern over asymmetry in personal info: should we require less info, and how would we deal with inaccurate notices?

 

Aistars: a lot of asymmetry. If you upload an image or © work, there’s not a lot asked of the user. No real attempt to educate about what’s appropriate or not. Some sites will point you to TOS.  On the flipside, takedown notices: most sites will walk you through the requirements; many of them will emphasize that there are penalties associated w/sending inaccurate notice and the fact that the notice will be forwarded to Chilling Effects and to the poster w/personal info.

 

Patrick Flaherty, Verizon: Verizon is also a © owner; we send DMCA notices, especially as we acquire content like AOL.  Notices are acted on quickly, and we don’t have a whack a mole problem. Our biggest concern is invalid notices related to P2P filesharing; Rightscorp notices make it harder for us to respond to legit notices and even crashed our server.

 

JC: a P2P notice isn’t legit?

 

Flaherty: there’s no takedown capability.

 

JC: does that mean you can’t receive a notice of infringement? I understand there’s no takedown piece, but forget about 512, what if I just sent a lawyer’s letter about illegal stuff flowing through your pipes?

 

Flaherty: that’s an allegation of infringement, but not a DMCA notice.

 

JC: if it’s formatted like a 512(c) notice, you take them in.

 

Flaherty: they arrive and we process them.

 

JC: why do you do that?

 

Flaherty: they come in by email when people use our DMCA form.

 

JC: but you don’t act on them.

 

Flaherty: We reject them.

 

JC: Do they know that?

 

Flaherty: It’s not automatic, but when someone gets to it, they get a response.

 

JC: So you just reject P2P allegations?

 

Flaherty: yes.

 

Isbell: do you keep track of those for repeat infringers?

 

Flaherty: not for conduit activity.

 

Isbell: you don’t have a repeat infringer policy?

 

Flaherty: we do, but not for these invalid notices.

 

Claggett: what would be a repeat infringer? Court decision?

 

Flaherty: we think courts are best determiners, though we do have voluntary agreements and forward notices under those.

 

Claggett: do you just consider them to be improper and think you can’t sue?

 

Flaherty: we consider them an abuse of the DMCA, but there’s no specific wording that points to them currently.

 

JC: For those who are looking for staydown, should that be focused on full length works? How do you deal w/potentially licensed use of second appearance?  Other related Q: Seems to be diversity of experience as between individual © owners and corporate systems.  Does anyone want to elaborate about ability of smaller © owners to access takedown tools.

 

Stephen Carlisle, Nova Southeastern: I run a © blog, and have one legacy client who’s a small © owner who controls about 50 composition.  I tried to send a takedown to Google, and if you go to Google’s © Office address—if you copy and paste, you don’t get to a takedown form, but several pages away from takedown. You’re asked to justify the takedown; if you say you’re the subject of the photograph, you get a warning saying that you’re probably not the © owner as if there were no such thing as selfies.  You have to create a Google account to file a takedown notice, which requires you to agree to venue in Google’s favor. Google creates these barriers.

 

Michael Michaud, Channel Awesome: Takedowns issued on things w/no accuracy—totally false, or reviews.  People that are trying to counter false takedowns, and they too have to give personal information; they are generally counternoticing against bully companies and they get scared.

 

Lisa Hammer: as small content provider I’d love more education/outreach; I’ve had trouble proving it’s my ©. 

 

JC: would it help to have a standardized universal takedown form.

 

Hammer: possibly; not an expert.

 

JC: One of the things in the comments was that there’s lots of variation w/service provider requirements. [b/c Usenet is not Google search is not Blogger!]

 

Hammer: sure, if it was in © Office.

 

Madaj: we don’t use automated processes.  We tried to use Google’s bulk tool, but the formatting process was difficult and time-consuming.  In addition, a number of people have said there’s a mechanism for expeditious removal, but there’s no clear definition of what’s required for expeditious removal; we’re forced to go back and check constantly. Sometimes it takes weeks, but we’re not clear on what the timeframe is to keep the safe harbor. 

 

JC: so you’ve had to wait weeks.

 

Madaj: yes.

 

Ellen Schrantz, Internet Ass’n: Staydown improperly shifts the burden. Endangers fair use and other limitations and exceptions; harm to free speech and creativity online. Educational efforts: our companies have multiple education efforts. The reason they can innovate and educate creators is b/c of the DMCA. If you endanger those protections, they may not be able to do that.

 

Maria Schneider: as it stands, the DMCA is the goose that laid the golden egg for these companies. They can do anything they want. YT requires you to get a Google account, accepting liability, TOS. It is such an intimidating process for a musician at every level.  YT allows you to upload a full album in 4 seconds. There’s no fair use.  They use these technologies against us to make it a constant battle. Staydown requires a fingerprint that they all have to use in the same way: Content ID or Audible Magic. Even Content ID, they’re actually using it to monetize, not to catch infringement.

 

Claggett: Has anyone challenged these requirements to sign up for an account or require additional info.

 

Schneider: I’m challenging it today!  The intimidation is great; musicians are scared to death. YT is offering $1 million to protect people. To Flaherty: what we need is a system to rate people; I’ve never had a counternotice against me.  It should come down immediately when I file.  I should have one-click access to Google.

 

Steven Rosenthal, McGraw-Hill Educ.: burdensome to monitor seemingly endless infringement on internet. Proliferation of educational content impacts our ability to ensure integrity of pedagogical process. Sites take offensive measures to prevent automated scraping by limiting metadata that we use. [McGraw-Hill’s sites: do they prevent automated scraping? Perhaps they have a reason?]

 

Melvin Gibbs: Flaherty’s comments showed massive asymmetry b/t small businesses and large ones.  [Hey, there are small ISPs too! I guess I’ll talk about those when I get a chance.]  We need to rely on good faith behavior.  We don’t have any voice on statutory licenses as musicians.  The idea that there will be collaboration is implausible. How I see this: the way the DMCA works now is de facto subsidy for large corporations. Our license rate is shrunken by asymmetry in our vulnerability and inability to enforce our basic rights.

 

George Johnson: Problem is that the burden of proof is on © creators; the burden should be on Google and the people who license our stuff. There has to be some fingerprint technology or metadata.  Burden of proof shouldn’t be on © creators.  [… who is it who posts on YT again?]  We can talk all day, but ultimately Congress can’t agree on anything, so we’re doomed.  As a creator, I can’t file a lawsuit against Verizon.  Law was written to benefit Google and licensors, not creators.  [Pause to note that DMCA predates Google by a few years.]

 

Eugene Mopsik: Rights holders are just looking for reasonable piece of the pie. For image creators there are automated services. But they come back w/100s or 1000s of occurrences of your image, and you have to evaluate whether any of those are licensed uses. Until there are persistent machine-actionable identifiers that aren’t easily scrubbed, and the Plus Registry is in effect and searching the web to add identifying info, we’re at a loss for having any control over images.  [And changing 512 would do all that, right.]

 

Alisa Coleman: You don’t have backend access to YT w/o having agreement w/them. Small companies don’t have the resources.

 

JC: third party vendors?

 

Coleman: cottage industry of takedown cowboys. Have to figure out who you want to be in bed with and who will represent your interests.  There are a lot of companies that don’t have takedown mechanisms w/ and we’ve negotiated the ability to whitelist or blacklist songs/recordings.  Catch-22/whack a mole still required.  They do have the ability to take things down and keep them down. They can make that happen. [Who they?]

 

Richard Burgess: there are a lot of companies that do takedowns, but there’s no reason the onus should fall on the © owner. The ISP companies create the problems; they make fortunes from the safe harbor and monetizing piracy.  Small or medium sized enterprise and individual © creators don’t stand a chance unless this is changed.  Second uses: so many tech solutions to that we could arrive at if we were in reasonable dialogue.  DMCA was written to provide a balance b/t service providers and creators, but that balance is gone.

 

Deborah Robinson: at Viacom we’re very sure about who owns exclusive licenses, so we don’t worry about second instances being different. 

 

Jonathan Band: Amazon would oppose staydown.  In addition to Schrantz’s reasons, it would be very hard for small startups to comply, and advantage big companies like Amazon.

 

JC: what if there were some consideration about size?

 

Band: that would address my issue but not others, like fair use. It’s great if voluntary arrangements exist, but the law shouldn’t be changed.

 

David Kaplan: Reiterate Robinson: we are sure who’s licensed.  Misunderstanding about online internet enforcement. Vendors don’t scan entire internet; focused on specific sites.  Use of tech to identify content is necessary for staydown. We use that widely in Content ID. Makes sense to focus on full length content as priority area.

 

Kathy Garmezy: ease of access is incredibly important.  Priority to full length clearly not fair use would be a starting point. Smaller companies need access to tech. Worth considering policy for repeat infringers.

 

Janice Pilch: one challenge for people w/o automated processes is identifying stuff in the first place, which is left to chance.  Currently rightsholders and creators w/o automated systems have to police constantly and forever.  The time has come to make effective standard technical measures available to anyone on reasonable and nondiscriminatory terms, make them open source, make ISPs responsible for monitoring and affirmatively seek facts consistent w/infringement.

 

Samantha Schonfeld, not speaking on behalf of Amplify Education: concerns of rights holders are no surprise to this group.  Past square peg; unpredictable future is round hole.   Echo Robinson & Pilch: this is a tech problem that lends itself to tech solutions; encourage panel to explore creative tech and legal solutions including w/o limitation compulsory license and compensation schemes.

 

Victoria Scheckler: we believe that there are tech solutions. Thoughtfully implemented but can address fair use.  Price: there are commercially reasonable available solutions.