I’m shaking my head at how many politicians block their constituents on social media. What a terrible practice. Today’s case is particularly mind-blowing. Rep. Reisch is a representative in the Missouri House of Representatives tweeting @CheriMO44. The court describes the allegations:
On June 22, 2018, Reisch tweeted about her appearance at a public event, at which her political opponent in the November 6, 2018 general election, Maren Jones, also appeared. Resich tweeted “Sad my opponent put her hands behind her back during the Pledge.” On June 23, 2018, Representative Kip Kendrick criticized Reisch’s tweet by commenting, “Maren’s father was a Lieutenant in the Army. Two of her brothers served in the military. I don’t question [Maren’s] patriotism. That’s a low blow and unacceptable from a member of the Boone County delegation.” Campbell retweeted Kendrick’s criticism in Reisch’s Account. After this retweet, Reisch permanently blocked Campbell from following or commenting on her Account.
FFS. Blocking a constituent because they retweeted a different representative (i.e., Reisch’s colleague)? Is this junior high?
Procedurally, the case presents itself as a motion to dismiss a 1983 claim. Substantively, the court says the following things:
“The parties do not dispute that Campbell’s re-tweet constituted speech protected by the First Amendment.”
“the fact that Twitter is privately owned does not preclude a finding that it is susceptible to the public doctrine analysis.” Cites to Packingham, Davison v. Loudoun, Davison v. Plowman, and Price v. NY.
“the public forum doctrine applies to Reisch’s Twitter Account. In particular, the Court finds convincing the rationale in Knight that the interactive space following each tweet in which other users may directly interact with the content of the tweets is subject to forum analysis.”
it doesn’t matter if the Twitter account is a public forum, limited public forum, or non-forum if the politician engaged in viewpoint discrimination.
“to the extent Reisch argues that she might reasonably decide that no Twitter account is better than one open to profane, racist, or obnoxious content, such viewpoint-neutral restrictions are not at issue in this case.”
“Campbell alleges that Reisch uses her Account under the handle of @CheriMO44 “to address her constituents, tout her accomplishments as a state representative, and promote her political agenda,” “to engage in political discourse related to her official duties as a state representative” and “open[ ] her account to comment by the general public.” In the Court’s view, these allegations are sufficient to plead that Reisch acted under the color of state law.”
“Campbell’s allegations create a reasonable inference that by blocking Campbell in response to his criticism of her tweet regarding her opponent following her appearance at a public event, Reisch was acting under the color of state law”
I don’t see these cases as a close call. If a politician uses a social media account to discuss anything remotely professionally-related–and almost every personal social media account will eventually cross over–then shutting down the ability of constituents to interact with the account is clearly censorship. Full stop.
As I’ve mentioned before, I think the social media providers could do more here. They should badge politicians’ accounts as official government channels and possibly restrict or eliminate those accounts’ ability to block constituents. Ultimately, maybe they should kick politicians off their social media services altogether to eliminate the inevitable disinformation campaigns that seem to follow politicians.
I don’t know if taxpayers are paying Reisch’s defense for this litigation. If they are, they should be angry that their elected politician is so thin-skinned as to block someone who retweets someone else’s post and their hard-earned tax dollars are being wasted to justify that. Even if Reisch is paying her own defense costs, I can’t imagine respecting any politician who intentionally shuts down a way for constituents to express their opinions. VOTERS: I ENCOURAGE YOU TO VOTE OUT ANYONE SO ILL-EQUIPPED FOR THE JOB OF REPRESENTING “THE PEOPLE.”
A pretty interesting copyright dispute is brewing in unexpected circumstances: among rival car washes in Fresno. The plaintiff includes the following language in its brochure and inspection sheet:
Dear Customer, While it is our intent to provide you with the best available service and the finest craftsmanship in the reconditioning of your vehicle, it has become necessary for us to disclaim total liability for any damage including but not limited to the following: engine damage, further chipping of the paint, tear to the interior fabric of worn material,…However, we have 100% confidence in the safety of our process…
The defendant copied this language verbatim for its own inspection sheet. I doubt it consulted a lawyer before doing so. Had it done so, I expect most copyright experts would have approved the recycling based on the merger doctrine and the Morrisey case involving the copying of sweepstakes rules. While it would be possible to reword the liability disclaimer better, there are only so many ways to express the underlying concepts that (a) we try to do a good job, but (b) we’re not liable for any damage, including [list of concerns], but (c) we’re confident our process is safe.
Nevertheless, the plaintiff got a copyright registration for its brochure and sued the defendant for copyright infringement. First, really? Second, this should be an easy defense win, right…?
The court rejects the defendant’s motion to dismiss:
the Court cannot conclude from the pleadings alone that merger or scènes à faire apply. First, Defendants’ conclusion that a liability disclaimer can be expressed in only a few ways is unsubstantiated. It is not readily apparent from the FAC that there are sufficiently few means of expressing the ideas in the disclaimer such that the ideas and expressions merge, or that the language of a disclaimer is akin, in the legal profession, to a stock scene or character. Defendants offer no authority suggesting that disclaimers of this sort can only be crafted in one or two ways. Instead, Defendants attempt to shift the burden of showing that there are many ways of expressing the idea of a liability disclaimer onto Plaintiff. Once the presumption of copyrightability has been established by the existence of a copyright registration certificate, it is the burden of the party challenging copyrightability to show that an exception applies. Moreover, on a motion to dismiss under Rule 12(b)(6) the Court assumes the truth of well pled allegations and does not engage in factual determinations. There is a marked and conspicuous similarity in the two disclaimers, including word choice and phrasing, but it is not readily apparent from the compliant that the similarities are owed to the nature of a disclaimer.
Even if the Copyrighted Material is a functional work, or merger or scènes à faire applied, it is still entitled to thin protection. The Court cannot conclude as a matter of law that thin protection would not bar the alleged infringement. A reasonable finder of fact could conclude that the disclaimer of liability contained within Defendants’ form is virtually identical to its opposite number in the Copyrighted Material. Notable is the identical wording in the salutation and first sentence, the first clause of which is non-functional as a disclaimer in that it expresses in “advertising speech” the intention to provide excellent service. The only changes, apart from those mandated by the layout, are the alteration of the word “best” to all-caps, the addition in Defendants’ disclaimer of a comma between “service” and “and,” the replacement of underlining with all-caps and the deletion of the word “any” in the phrase “necessary for us to disclaim any and all damages,” and omissions to the list of specific types of damage for which liability is disclaimed. Defendants’ sheet also includes a bolded statement to the effect that car wash employees are not permitted to remove child seats or other child restraints. Therefore, even assuming that the Copyrighted Material is entitled only to “thin” copyright protection, Defendants’ motion to dismiss fails.
Ugh. There are some obvious problems here, including (1) the merger doctrine doesn’t require the defendant show that there are only “1 or 2” ways of expressing the idea; and (2) a work subject to the merger doctrine isn’t protected by a thin copyright, it loses its copyright protection altogether. I sure hope defense counsel helps the judge better under the merger doctrine sooner rather than later.
This case is also a good example of the problems with weaponizing thinly copyrighted works. The work in this case is essentially worthless as a copyright artifact, yet it will do a lot of damage before it’s dead.
More conceptually, the judge seems to valorize the copyright over the obvious competition problems lurking in this case. (Cf. one of my least favorite copyright cases of all time, the Webloyalty case). The brochure doesn’t cause any real competitive injury; so making this a federal copyright case is a huge waste of time and resources. In fact, if there’s any competitive injury here, it’s due to the maintenance of this dubious lawsuit designed to squelch competition. Perhaps the judge will address this concern later in the case, but he could have done everyone a favor by squashing this case early rather than letting it drain everyone’s coffers.
Yesterday, I did a webinar for the California Lawyers Association on the status of the California Consumer Privacy Act (CCPA). This post recaps the discussion.
A Quick Overview of the CCPA
The CCPA imposes 6 new obligations on covered businesses: they have to make specified disclosures to consumers, provide consumers with a data erasure capacity, provide consumers with data portability, allow consumers to opt-out of data sales (or opt-in in the case of minors), and not discriminate against consumers on the basis of personal information. The law also creates a private cause of action for certain data breaches.
The CCPA applies to any “information that identifies, relates to, describes, is capable of being associated with, or could reasonably be linked, directly or indirectly, with a particular consumer or household.” It’s my position that this includes any data about consumers, so this definition effectively eliminates the notion of anonymized or unattributed data.
The CCPA regulates every business with $25M+ in annual revenues, or 50%+ of their revenues from selling consumer data, or that “annually buys, receives for the business’ commercial purposes, sells, or shares for commercial purposes, alone or in combination, the personal information of 50,000 or more consumers, households, or devices.” The CCPA purports to apply unless every aspect of commercial conduct takes place wholly outside of California.
The CCPA affects many businesses who never had a chance to explain the law’s problems to the legislature;
The CCPA imposes excessive costs on small businesses;
The CCPA requires businesses to waste money complying with multiple privacy laws;
The CCPA degrades consumer privacy in several ways;
The CCPA’s definitions are riddled with problems; and
The CCPA reaches beyond California’s borders.
Another big problem not mentioned in the letter: the CCPA was dumped on California’s extensive tableau of existing privacy laws with minimal efforts at harmonization. Almost certainly, some of the pre-existing laws need to be deprecated; and other laws will not make sense unless/until they are conformed. I’m not sure who will do that harmonization work.
In preparation for their rule-making process, the AG’s office has held 6 “listening sessions,” with one more remaining, in the following places: January 8, San Francisco / January 14, San Diego / January 24, Inland Empire / January 25, Los Angeles / February 5, Sacramento / February 13, Fresno / March 5, Palo Alto.
Each listening session starts with the AG’s office presenting this slide deck. Then, audience members get a chance to speak up. Many speakers just read what they’ve already submitted to the AG’s office in writing. There is no reason to believe that the AG’s office will give more weight to oral comments over written submissions, so there is no reason to attend the hearings or get hung up on recaps of the hearings. Still, I’ll be at the Palo Alto session if anyone wants to have a privacy party.
Written comments to the AG’s office are due March 8. This is absolutely the best time to get your comments into the AG’s office–before they have put pen to paper.
Two big questions:
1) Will the AG’s office make any rules addressing provisions beyond those expressly delegated to them in the law? The CCPA enumerates the following specific areas for AG rulemaking:
(1) Categories of Personal Information
(2) Definition of Unique Identifiers
(3) Exceptions to CCPA
(4) Submitting and Complying with Requests
(5) Uniform Opt‐Out Logo/Button
(6) Notices and Information to Consumer, including Financial Incentive Offerings
(7) Verification of Consumer’s Request
Very few public comments have addressed any of these topics, because most of these issues are far less problematic than the CCPA’s big structural problems. So what will the AG’s office do? Will they restrict their rule-making to the explicit sandbox designated by the legislature, or will they respond to the comments and spread their wings? I fear there will be some diffusion of responsibility: the legislature will think the AG’s office can make rules about whatever they want (that’s literally what the law says), and the AG’s office will feel constrained to defer to the legislature and will only do what the law expressly directs them to do.
2) When will the rulemaking be done, and when will the law go into effect? The AG’s office put together this handy chart of their process:
As you can see, the AG’s office plans to issue the first draft of their rules in “Fall” 2019. The draft rules will almost certainly generate a massive number of comments, plus there will be scheduled hearings to discuss them. If the first draft of the rules is issued Sept. 21, 2019 (the first day of Fall), there is absolutely no way the AG can finalize the rules before January 1, 2020. The law takes effect 6 months after the AG’s final rules, with a hard stop of July 1, 2020. I think it’s guaranteed that the law’s effective date will be July 1, 2020 and that businesses will have less than 6 months–possibly way less–to adopt their practices to the AG’s rules. It would be appropriate for the legislature to slip the July 1, 2020 hard stop further, but I doubt that will happen because the legislators will demand too high a price from “business” for this “concession.”
Developments in the California Legislature
There will be many bills to amend the CCPA. I don’t yet how many. Here is an incomplete roundup.
Looking at the broader context, there are few business allies in Sacramento. Support for improving CCPA will not come from the governor’s office: Gov. Newsom used his State of the State address to endorse CCPA and propose a “data dividend” (presumably, just another tax) from Internet companies. It doesn’t appear oversight will come from the Republicans either (not that they have a lot of clout in the legislature any more); several Republicans already introduced a privacy bill (AB288, the Own Your Own Data Act) to extend, not curtail, CCPA.
I don’t see the California legislature having much appetite to fix the CCPA, either. There will be a torrent of post-CCPA privacy regulatory bills soon introduced into the California legislature. As those new bills flood the legislature, it will dilute the time and focus the legislature has to fix the many problems with the CCPA. CCPA is already old news to the legislature, while it’s a lot more exciting to manufacture brand new laws. So the odds of getting serious reform to the CCPA will be undermined by the fervor to pass more privacy laws.
Several states have introduced bills to adopt variants of the CCPA, including Washington states. Bloomberg reports on the efforts in 8 states. This proliferation is bad news. Either the other states’ laws will fix the CCPA’s mistakes (or possibly add their own) and thus increase divergence among state laws, or the bills will copy California’s mistakes and further lock in terrible policy. Either way, no one wins.
There have been numerous federal privacy bills introduced. Some propose to preempt state laws; others do not. I’ve said it many times that the sine qua non of a new federal privacy law is state law preemption. Otherwise, it just adds more law and more complexity without fixing any of the problems in states’ divergent or misguided laws (cf. how the DTSA didn’t clean up trade secret litigation, it just proliferated plaintiffs’ claims). From my perspective, a federal preemptive law is the only remaining hope we have to avoid overregulation of privacy that will destroy the Internet and put a serious dent in our economy. Yet that slim hope relies on a dysfunctional Congress, which isn’t reason for optimism.
There have been murmurs about Constitutional challenges to the CCPA. There are several angles that could be meritorious. However, I have yet to hear of anyone take any demonstrable steps forward in bringing these challenges.
I did an interview with Vanessa Blum of the Recorder about emoji law. It was first published here. The text:
* * *
Santa Clara University law professor Eric Goldman is an expert on internet speech. Lately, he’s been rather obsessed with a topic that most lawyers probably have never stopped to think about—the ways that emojis are creeping into court cases and how judges and litigants are dealing with them.
Emojis can be difficult to interpret when they’re presented as evidence in a case, and that issue isn’t going anywhere in the near future. We checked in with Goldman to see what’s new in emoji law. Here’s what he had to say:
What areas of law are being impacted by emojis—and can you give a few quick examples?
Goldman: Emojis show up in virtually every practice area because emojis are showing up across all types of online communications. Emojis show up most frequently in cases where online chatter is a key source of evidence. By far the most common types of cases involving emojis are sexual predation cases, where the perpetrator and the victim (or an undercover cop) exchange flirtatious or sexual banter, often a form of victim grooming. For similar reasons, employment discrimination cases show up disproportionately frequently in my list of emoji cases.
In 2017, you wrote a paper that surveyed emoji references in about 80 court opinions and warned lawyers and judges to prepare for “the coming emoji onslaught.” What are some more recent developments that have come to your attention?
Unfortunately, opinions still struggle with displaying emojis. Opinions routinely omit the emojis altogether, or the judge imprecisely characterizes the emoji(s) in evidence. (It doesn’t help to call an emoji a “smiley” because there are a dozen different smiling emoji symbols). Furthermore, Westlaw and Lexis still usually do not display emojis, and neither database makes it possible to search for emojis in court opinions. So while the number of court opinions referencing emojis is growing rapidly, the court publication process remains woefully under-prepared.
Another recent development: In 2016 and 2017, there were still substantial differences in how online platforms depicted the same emoji symbols. In the past couple of years, many platforms have done a lot to harmonize their depictions of emojis. The depictions between platforms are not as starkly different as they used to be. However, there are still inexplicable minor variations across platforms, so real harmonization of emoji depictions seems far away.
The diversity of emoji depictions across platforms will inevitably cause lawsuits. As one empirical study recently found: “at least 25% of respondents were unaware that the emoji they posted could appear differently to their followers. Additionally, after being shown how one of their tweets rendered across platforms, 20% of respondents reported that they would have edited or not sent the tweet. These statistics reflect millions of potentially regretful tweets shared per day because people cannot see emoji rendering differences across platforms.” Those “regretful tweets” should make lawyers some good money.
Isn’t actual language also subject to varying interpretation? Are emojis extra challenging in this regard or just another way that human communication is often blurry?
Goldman: You’re absolutely right. Courts have been interpreting nonverbal/non-textual communications for centuries. Indeed, interpreting communications between people is one of the strengths of our judicial system. In that respect, emojis are just another type of nonverbal/non-textual communications for courts to interpret.
However, emojis do have some unique attributes that require extra consideration when interpreted. I’ll mention just three key differences (there are more).
First, emojis often are visually quite small, and there are many emoji symbols that look similar. Therefore, they pose a greater risk that the reader will incorrectly, but reasonably, decode an emoji symbol compared to many other forms of communication.
Second, every type of communication develops regional and community-specific dialects. However, because emojis look different on different platforms, the symbols can develop platform-specific dialects. For example, on Apple, eggplants are associated with penises and peaches with butts because of how Apple depicted those emojis. For a long time, emoji users on other platforms had no idea of those connotations, because the eggplant and peach emojis on their platforms didn’t suggest those associations.
Finally, because emojis display differently on different platforms, it’s entirely possible for an emoji sender and recipient to see significantly different depictions of the same symbol, meaning that both may reasonably reach entirely different interpretations. For example, for a while, Google users thought the “grinning face with smiling eyes” emoji meant “blissfully happy,” while Apple users thought it meant “ready to fight.” Thus, a Google user sending that emoji symbol to an Apple recipient might unintentionally prompt violence, even though both the Apple and Google users made reasonable interpretations of the symbol. Determining how to apportion liability when both parties made reasonable but different interpretations of the same symbol will lead to many unhappy outcomes.
What is your advice for judges who confront emojis in their cases?
Goldman: I have three suggestions. First, judges should make sure that the lawyers present the exact depictions that their clients saw. There are so many circumstances where the sender and recipient saw different symbols, and the differences could affect the dispute. It would be a potentially major mistake for a judge or litigants to assume that there is a single canonical depiction of emojis that both parties saw identically.
Second, judges should make sure that the fact-finder gets to see the actual emojis so it can figure out its meaning directly. If testimony is being read in court, the emojis should not be orally characterized but should be displayed to the fact-finder.
Third, judges should display the actual emojis in their court opinions. They should not omit the emoji symbols or try to characterize them textually. The symbols might not render properly in print or Westlaw/Lexis, but they should at least appear properly in the PDFs of the court opinions.
I imagine that your writing on emojis brought attention to the issue before it had even occurred to most lawyers. Is there anything new you’ve started to see with online communication that is likely to start presenting issues in the courts?
Goldman: As unsettled as emoji law is today, it’s only going to get more complicated as technology evolves. For example, we generally assume emojis are static, but they can be animated. Apple’s animojis reflect the sender’s facial movements, and Samsung’s “AR emojis” are personalized animated avatars. These innovations pose even more challenges for courts to interpret, including the standard challenges associated with video evidence. Furthermore, because the unique attributes of these next-gen emojis are platform-specific, lawyers are going to need to be quite familiar with the platform’s operation to fully evaluate their cases.
As I’ve mentioned before, I track every U.S. court opinion in Westlaw and Lexis that references “emoji” or “emoticon.” This is not a comprehensive census for several reasons, including my inability to set up alerts when a court displays the symbol without calling it an emoji or emoticon (which, in many emoji cases, aren’t even displayed in Westlaw or Lexis) and the other known skews and limits of Westlaw’s and Lexis’ case collections. Still, FWIW, I’ve posted the updated roster of cases.
As expected, the number of emoji/emoticon case references are growing in a typical exponential J curve. By my count, there were 53 such cases in 2018, compared to 33 in 2017. I count a total of 171 cases all-time, of which over 30% were in 2018.
As the above chart shows, the relative frequency of the term “emoji” has completely eclipsed the term “emoticon.” The chart is skewed a bit by the fact some opinions use both terms, and I counted those opinions only as referencing emojis. Still, while emoticons are likely to remain part of the lexicon for the foreseeable future, clearly emojis are the future.
While the number of opinions referencing emojis is growing rapidly, 2018 did not see any major substantive rulings interpreting emojis. In 37 of the 53 cases, the term emoji or emoticon only appeared a single time, usually meaning that the emoji/emoticon was just an incidental part of the evidence. None of the other 16 cases broke any interesting new ground on emoji interpretations.
at least 25% of respondents were unaware that the emoji they posted could appear differently to their followers. Additionally, after being shown how one of their tweets rendered across platforms, 20% of respondents reported that they would have edited or not sent the tweet. These statistics reflect millions of potentially regretful tweets shared per day because people cannot see emoji rendering differences across platforms
When the authors say “millions of potentially regretful tweets shared per day,” the law professor translates that into “lawsuits aplenty.”
Finally, if you’re in the Bay Area on February 12 at noon, I’m speaking at Santa Clara Law about Emoji Law at the kickoff event for the law school’s brand new Internet Law Student Organization (check out our group photo). It’s open to the public, so we’d love to have you join us. Yes, I will be wearing my emoji tie at the talk!
Dreamstime sells stock photos. It started running Google AdWords in 2004. In 2015, Google organic (?) referrals to Dreamstime “plummeted,” allegedly reducing its number of new customers by 30% in a year. (The opinion isn’t crystal-clear that the traffic drop was only organic). In response, Dreamstime took many of the standard steps to rehabilitate its search reputation, plus it increased its AdWords spend by 50%. According to the complaint, “no one, including ‘guru’ search engine optimization companies, have identified any issues with Dreamstime’s site that would explain how Dreamstime came to be marginalized in Google’s search rankings.” Meanwhile, other search engines have been kinder to Dreamstime. For the search term “stock image,” Dreamstime claims it ranks 91st on Google (lower than some clearly marginal sites), 5th on Bing, 4th on Yahoo, and 3rd on Baidu. Dreamstime also alleges various Google chicanery about the AdWords ads. Dreamstime claims that Google deliberately hurt it while Google favored other competitors that were in partnership with Google. As a result, Dreamstime brought an antitrust suit against Google.
Sherman Act Monopoly Maintenance. Dreamstime doesn’t allege that it competes directly with Google for advertiser dollars or that Google leveraged its monopoly position. Instead, it apparently alleges that Google was illogically mean to a paying customer. The court says those allegations do not show any harm to competition; “by destroying Dreamstime, no rival and no competition has been excluded from the online search advertising market, and therefore, no anticompetitive conduct has been adequately alleged.”
Dreamstime also complained about Google’s data collection, and that goes nowhere: “Although the data collection likely gives Google an advantage in the online search advertising market over its rivals, a monopolist utilizing its competitive advantage does not equate to anticompetitive conduct.” The court continues:
A company providing a platform for businesses to sell advertisements must be expected to efficiently produce the best possible product to its customers. Thus, if in the online search advertising market, this means making stock photographs readily available to businesses seeking to advertise, a company will likely attempt to enter into agreements to make those stock photographs as available as possible. The use of data and the securing of these agreements here are efficient business transactions that do not “attempt to exclude rivals on some basis other than efficiency.”
The court dismisses the antitrust claim with prejudice. However, in an unusual twist, it says Dreamstime’s discovery requests for its surviving claims can also apply to this (dismissed) claim, and I believe the court says Dreamstime can ask for permission to resurrect the antitrust claim if it finds anything good in discovery.
Contract Breach. The court says Dreamstime properly alleged that Google overdelivered AdWords clicks in excess of requested limits and improperly deindexed its app in Google Play.
Implied Covenant of Good Faith and Fair Dealing. The court says these allegations sufficed:
Google pretended to work to resolve purported “policy issues” with Dreamstime’s ads while actually subverting and frustrating the ability of Dreamstime to realize the benefits of its contract. Google also is alleged to apply a double standard to Dreamstime’s low cost and high cost advertising campaigns and a double standard in Google’s treatment of Dreamstime compared to Dreamstime’s competitors.
The 17200 claim also survived.
Implications. Despite the favorable ruling on the Sherman Act claim, this is surely a frustrating ruling for Google. Dreamstime is almost certainly going to lose this case eventually. However, by greenlighting the case to the next stage and opening up discovery–including investigations for a dismissed claim–the judge ensured that the parties will spend a lot of time and money reaching the inevitable denouement. Even if the judge technically got the legal standard right for a motion to dismiss, this is one of those situations where Google surely wished the judge would demand more rigorous factual support in the complaint. I’m confident this judge will eviscerate Dreamstime’s case when it fails to deliver on its allegations with credible and admissible evidence, but it will take a lot of resources from both parties to prove that negative.
This post rounds up some FOSTA-related links I’ve aggregated over the past few months. There is no good news here. The data points suggest that in FOSTA’s first 9 months, it apparently has failed all of its policy goals while causing substantial collateral harm. Some of the lowlights (with citations):
Online sex ads remain prevalent after FOSTA and the Backpage shutdown.
AP: Online sex ads rebound, months after shutdown of Backpage
FOSTA has increased the number of sex workers on the street, where they face substantial physical harm and the control of pimps.
AP: Side effect of trafficking law: More street prostitution?
SF Chronicle: The Scanner: Sex workers returned to SF streets after Backpage.com shut down:
“Without being able to advertise online,” Long said, “a huge number of sex workers were forced to go outside, and many have reported that former pimps came out of the woodwork offering to ‘manage’ their business again since they were now rendered unable to find and screen clients online.”
St. James saw a spike in street-based sex work in the month after Backpage.com was shut down, Long said, and screening clients has become more difficult because of the closure of other sites popular in the industry.
“The very bill that was supposed to stop trafficking has quite literally given formerly irrelevant traffickers new life,” Long said.
Law enforcement has virtually stopped looking for sex trafficking victims online. Instead, law enforcement has redeployed its resources towards busting sex workers on the streets. This does nothing to help sex trafficking victims and further victimizes sex workers.
The Internet has shrunk in important ways. The shutdown of Craigslist personals was the most dramatic shrinkage, but it was hardly the only one. In what are almost certainly FOSTA-related moves, Patreon and Tumblr reduced their support for adult content and Facebook adopted a silly policy regarding profanity. Misogynists are also pursuing adult online content creators using takedown notices; and those notices are getting extra attention from online services due to FOSTA.
The Stranger: “Tumblr banning porn, adult content, & ‘female-presenting nipples’…serves as a grim portent for what FOSTA-SESTA will do to the internet as we know it….It’s a canary in the coal mine, warning of a more censored, puritanical internet”
Wired: Misogynists are attacking female pornography producers & sex workers by weaponizing the online payment systems’ overresponsiveness to unsubstantiated user reports
Techdirt: After Getting FOSTA Turned Into Law, Facebook Tells Its Users To Stop Using Naughty Words
Techdirt: NY Times Columnist Nick Kristof Led The Charge To Get Facebook To Censor Content, Now Whining That Facebook Censors His Content
After FOSTA’s enactment, the DOJ busted another online “escort” site–ONCE AGAIN using older laws rather than the new provisions created by FOSTA.
DOJ: “Manhattan U.S. Attorney Announces Money Laundering Charges Against Operators Of Nationwide Prostitution Enterprise And Seizure Of Online Escort Website.” The DOJ busts another online service, “Flawless Escorts,” based on money laundering, not FOSTA or the SAVE Act.
Facebook is being sued in Texas state court for facilitating sex trafficking, but the plaintiffs’ claims don’t allege new FOSTA claims either.
Basically…WTF? Where is the win–any win–here for anyone? All FOSTA supporters should be alarmed by what they are seeing–and taking a hard look in the mirror.
Bonus: updates on the Backpage criminal prosecutions:
* Reason: The Senate Accused Them of Selling Kids for Sex. The FBI Raided Their Homes. Backpage.com’s Founders Speak for the First Time.
* Law.com: Delaware AG Sues to Shut Down LLCs Linked to Backpage.com Sex Trafficking. The complaint.
Six Flags, the amusement park operator, allegedly violated Illinois Biometric Privacy Statute by collecting a minor’s fingerprint without consent. The state appeals court held that plaintiff had to allege harm beyond the information collection without consent. The Illinois state supreme court reverses.
As alleged in the complaint, Alexander Rosenbach went on a field trip to Six Flags. His mom purchased a pass for him online and expected him to take care of the paperwork on the trip. Alexander was fingerprinted at Six Flags before he was issued a season pass. Upon his return, his mom asked him for the paperwork accompanying the pass, and he said it was all taken care of “by fingerprint now.” She sued on his behalf, alleging a violation of the Illinois biometric privacy act (BIPA). This statute requires disclosure and consent before biometric information is collected.
The court says that the Illinois legislature knows how to make clear their intent to limit standing to plaintiffs who suffered “actual damage” beyond a violation of rights bestowed by statute. The legislature did not reflect such an intent here. This biometric privacy statute is similar to the AIDS Confidentiality Act, another Illinois statute where the legislature said plaintiffs can sue by showing any violation of the statute.
The statute says anyone “aggrieved” may sue and doesn’t define the term further, so the court looks to its commonly accepted definition. A person is aggrieved when her “rights are invaded . . . or . . . pecuniary interest” adversely affected. This definition has been recognized numerous times by Illinois courts, and the legislature is presumed to know this definition.
The court also says that requiring additional harm beyond a statutory violation is at odds with the purpose of the statute, which gives individuals the right to control their biometric information by requiring notice and consent before collection. Collecting this information without consent frustrates this purpose:
This is no mere ‘technicality’. The injury is real and significant.
The court also cites to the growing trend of collection and use by companies of biometric information. The court also notes that the only enforcement mechanism in the statute is a private right of action. And compliance “should not be difficult.”
The Illinois BIPA statute has become the bane of tech companies. The court also notes a Facebook case that reached a similar conclusion. I blogged the trial court’s denial of Facebook’s summary judgment ruling in that case. Since that blog post, the Ninth Circuit granted Facebook’s request for discretionary appeal, staving off trial for the moment.
It appears from the pleadings that Six Flags did not have any documentation at all relating to the season pass. The opinion alludes to an online purchase, but there is no discussion of any accompanying terms. One wonders what the outcome would be if users had an online sign up with a click-through disclosure and consent. Setting aside capacity to contract for minors, is that sufficient to inform the subject and constitute the “written consent” that is required? Privacy statutes often require consent, and to my knowledge no case has really dug into whether consent in a terms of service (even one that the person affirmatively consents to) is sufficient.
Eric’s Comments: As Venkat indicates, BIPA was passed in a different technological era, and it now casts a long and not-always-welcome shadow over a wide range of technological innovations. That could indicate that the BIPA drafters and Illinois legislature did a good job foreseeing the future, or it could indicate that BIPA was premature, drafted before we really understood biometric data and the potential social benefits–and challenges–of using that data.
Unless some of the other pending federal courts erase any Article III limits, this ruling virtually ensures that future BIPA litigation will likely take place exclusively in Illinois state court. Why would a plaintiff want to risk an Article III dismissal in federal court when it’s completely bypassable?
The court casually makes an unsupported factual claim that it would have been easy for Six Flags to get consent. I’d love to stress-test this assumption. Why is Six Flags relying on biometric data in the first place? Maybe it’s essential, or maybe it’s a frivolous convenience. How would Six Flags get consent from both the child and parent in this situation? Why didn’t they try? We know that in the photo facial recognition cases, it’s effectively impossible for the online services to get consent from third parties depicted in the images who aren’t in privity with the online services. If that fact had been before the Illinois Supreme Court, would it have still claimed that getting consent is easy?
Glassdoor allows employees to review employers. An employer, Andra, was unhappy about ten pseudonymous reviews posted about it from June 2014 to June 2015. In August 2015, it sought pre-suit discovery of the reviewers’ identities. Glassdoor opposed the request and filed a Texas anti-SLAPP motion to strike. In February 2016, the trial court denied the motion to strike and granted Andra’s request to unmask two of the 10 reviewers. The appellate court affirmed.
However, this case has been tied up in court long enough that Andra’s underlying claims against the reviewers are now barred by the statute of limitations. As a result, the Texas Supreme Court dismisses the case as moot.
The statute of limitations (SOL) calculation depends on the application of the single publication rule, something I haven’t blogged about in a while. Online, the SOL can start on the first upload, the last download, or something in between. Measuring the SOL by the last download constantly pushes back the date, often eliminating the SOL in practice. As a result, almost all of the courts addressing the issue start the SOL upon the initial upload. (This might be reset if the uploaded content is edited and the edits are relevant to the claim; that’s not an issue here). So the SOL for the 10 reviews started no later than June 2015, though the court slips the date to August 2015, when Andra demonstrably had discovered the reviews because it started the pre-suit discovery process then. Andra’s claims have a 1 or 2 year SOL, so all SOLs ran out no later than August 2017. Now, in 2019, the deadlines are passed.
To get around the SOL, Andra argued that “Glassdoor has absolute and complete ability to control and has undisputed title and interest in the posts.” Putting aside the dubious factual predicates of the claim, the court says that workaround argument has been tried and rejected before.
Andra argued that the rule should be different for “restricted access sites” like Glassdoor. The court rejects the factual claim: “posts to Glassdoor’s website are not confidential and that Glassdoor does not charge users for access. Once posted, the reviews are equally accessible to all users.” In a footnote, the court acknowledges (without endorsing) the argument that SOLs should slip when Internet publications are obscure, i.e., they can’t be viewed by the general public or aren’t indexed in search engines, but says it “need not address this concern here, as Andra highlights Glassdoor’s ‘high number of visitors and visibility to the general public.'”
Andra also argued that the reviewers’ pseudonymity meant that the SOL should be tolled until it discovered their identities. The court says that Andra could have filed a “John Doe” lawsuit and sought discovery; but instead it chose to seek pre-suit discovery against Glassdoor.
Unfortunately for Glassdoor, the mootness ruling also takes out its anti-SLAPP motion and the associated attorneys’ fee shift. Nevertheless, Glassdoor once again successfully defended its reviewers’ privacy, a result we all benefit from.
So here’s the dilemma that’s been bothering all morning. Texas has a pre-suit discovery proceeding to help plaintiffs bring their claims and possibly reduce the total amount of litigation. This ruling effectively punishes Andra for trying to take advantage of it, creating a procedural “gotcha.” When Glassdoor or its reviewers fight back, they can tie up the pre-suit discovery request in court long enough for the SOLs to run. So either Andra should skip the pre-suit discovery proceeding altogether and just start with a John Doe lawsuit, or if the proceeding gets tied up in court, it should bring a separate John Doe lawsuit before the SOLs run out. The obvious solution would be to toll the SOLs while the pre-suit discovery proceeding is pending, but the Texas Supreme Court expressly rejects that outcome. So I’m not sure if this ruling completely guts the pre-suit discovery process, but at minimum it makes the process less attractive and more risky to plaintiffs.
[This is part 2 of a 2-part series on the Music Modernization Act. Check out part 1 before reading this post.]
The Musical Works Database
One of the most important provisions in the MWM Act is the provision requiring the Collective to “establish and maintain a database containing information relating to musical works (and shares of such works) and, to the extent known, the identity and location of the copyright owners of such works (and shares thereof) and the sound recordings in which the musical works are embodied.” [17 U.S.C. §115(d)(3)(E)(i)] The database shall include the title of the work, the copyright owner(s) and ownership percentage(s), contact information for the copyright owner (if known), and “to the extent reasonably available,” “the international standard musical work code” for the work, and “identifying information for sound recordings in which the musical work is embodied,” including titles, featured artists, sound recording copyright owners, producers, and international standard recording code. [17 U.S.C. §115(d)(3)(E)(ii)] “[T]o the extent practicable,” musical work copyright owners have an obligation to “engage in commercially reasonable efforts” to supply such information to the Collective. [17 U.S.C. §115(d)(3)(E)(iv)] The statute also contemplates that there will be “unmatched works,” or shares of works, for which a copyright owner cannot be identified or located (i.e., orphan works), for which the database will include as much of the information as is known. [17 U.S.C. §115(d)(3)(E)(iii)] Most importantly, “[t]he musical works database shall be made available to members of the public in a searchable, online format, free of charge.” [17 U.S.C. §115(d)(3)(E)(v)] The database must also be made available “in a bulk, machine-readable format, through a widely available software application,” free of charge to digital music providers, significant nonblanket licensees, authorized vendors of the above, and the Register of Copyrights; and “for a fee not to exceed the marginal cost of providing the database” to “any other person or entity.” [Id.]
The Collective is also required to “maintain a current, publicly accessible list of blanket licenses” and another list of “notices of nonblanket activity” of significant nonblanket licensees, including contact information for both types of licensees and effective dates. [17 U.S.C. §115(d)(3)(F)]
Collecting and Distributing Royalties
Digital music providers must report and pay royalties to the Collective each month, 45 calendar days after the end of the monthly reporting period. [17 U.S.C. §115(d)(4)(A)(i)] The reports “shall be in a machine-readable format that is compatible with the information technology systems” of the Collective, and must include usage data for musical works under the blanket license, and for voluntary licenses and individual download licenses, along with identifying information for each sound recording and information concerning the authorship and ownership of each musical work. [17 U.S.C. §115(d)(4)(A)(ii),(iii)] The provider “shall engage in good faith, commercially reasonable efforts” to obtain this information from sound-recording copyright owners. [17 U.S.C. §115(d)(4)(B)]
Upon receiving “reports of usage and payments of royalties from digital music providers for covered activities,” the Collective has five obligations: (I)(aa) to “engage in efforts to” identify the musical works and their copyright owners, (bb) to confirm uses subject to voluntary licenses and individual download licenses, and calculate the amounts to be deducted from the royalties otherwise due under the blanket license, and (cc) to confirm payment of royalties due; (II) to distribute royalties to copyright owners that can be identified and located; and (III) to deposit unclaimed or disputed royalties into an interest-bearing account. [17 U.S.C. §115(d)(3)(G)(i)] Any additional royalties that are collected as a result of enforcement efforts shall be distributed in the same manner on a pro rata basis. [17 U.S.C. §115(d)(3)(G)(ii)]
A digital music provider is in default if it fails to file a monthly report, to make a monthly royalty payment or late fee payment when due, or to pay its administrative assessment, or if it provides a monthly report that is “materially deficient as a result of inaccurate, missing, or unreadable data” that was available to it; or if after being provided written notice, the provider refuses for at least 60 days to comply with any other material term. [17 U.S.C. §115(d)(4)(E)(i)] The Collective must then send written notice “describing with reasonable particularity the default,” and giving the provider 60 days to cure the default. [17 U.S.C. §115(d)(4)(E)(ii)(I)] If the provider fails to cure the default within 60 days, the blanket license automatically terminates, leaving the provider liable to an infringement action. [17 U.S.C. §115(d)(4)(E)(ii)(II)] The provider may seek judicial review of any such termination in U.S. District Court. [17 U.S.C. §115(d)(4)(E)(iv)]
No more than once every three years, the Collective may audit the records of a digital music provider to verify the accuracy of its royalty payments for the preceding three years. [17 U.S.C. §115(d)(4)(D)(I)] If there is an underpayment of ten percent or more, the digital music provider must pay the reasonable costs of the audit as well as paying the underpaid royalties. [17 U.S.C. §115(d)(4)(D)(VI)] The statute of limitations for any action to collect underpaid royalties is 6 years after the audit is commenced. [17 U.S.C. §115(d)(4)(D)(VII)]
If the musical work copyright owner cannot be identified or located, the Collective is required to hold the royalties in an interest-bearing account for at least three years after the royalties were accrued or received (whichever period expires sooner). [17 U.S.C. §115(d)(3)(H)] (Exactly how royalties could be received before they accrued is left unexplained.) If the musical work copyright owner is identified and located during that three-year period, the Collective shall pay the accrued royalties and a proportionate share of the interest to the copyright owner. [17 U.S.C. §115(d)(3)(I)]
The Collective must “maintain a publicly accessible online facility … that lists unmatched musical works (and shares of works)” and “engage in diligent, good-faith efforts to publicize … the ability [and procedures] to claim unclaimed accrued royalties for unmatched musical works.” [17 U.S.C. §115(d)(3)(J)(iii)] The Collective must also establish policies and procedures “to address in a timely and equitable manner disputes relating to ownership interests in musical works … and allocation and distribution of royalties.” [17 U.S.C. §115(d)(3)(K)(iii)]
If royalties for an “unmatched” work remain unclaimed after three years, the Collective shall distribute the unclaimed royalties and interest to other musical work copyright owners, “in a transparent and equitable manner based on data indicating the relative market shares of such copyright owners.” [17 U.S.C. §115(d)(3)(J)(i)(II)] (Market shares are to be based on usage data reported to the Collective, but the Collective “shall take appropriate steps to safeguard the confidentiality and security of” the data. Id. The Act does not specify how the Collective is to harmonize “transparency” with “confidentiality.”) “The first such distribution shall occur on or after January 1 of the second full calendar year to commence after the license availability date.” [17 U.S.C. §115(d)(3)(J)(i)(I)] The license availability date is January 1, 2021, so the date specified to begin distributions is “on or after” January 1, 2023. This creates a logical conundrum, since reports and royalties are only due beginning on the license availability date, so the three-year waiting period cannot have expired before January 1, 2024. One suspects this is simply a typographical error, and that Congress will likely be asked to fix it sometime before then. In any case, subsequent distributions must take place at least once every calendar year after that. [17 U.S.C. §115(d)(3)(J)(i)(I)] Musical work copyright owners must then pay at least 50 percent of those distributions to songwriters, in proportion to the reported usage of their musical works. [17 U.S.C. §115(d)(3)(J)(iv)]
The Collective is not liable for “good-faith administration of policies and procedures adopted and implemented to carry out” these provisions, “except to the extent of correcting an underpayment or overpayment of royalties.” “Good-faith administration” means “in a manner that is not grossly negligent.” [17 U.S.C. §115(d)(11)(D)] However, “the collective may participate in a legal proceeding as a stakeholder party if the collective is holding funds that are the subject of a dispute between copyright owners.” [Id.] “The holding and distribution of funds by the [Collective] … shall supersede and preempt any State law … concerning escheatment or abandoned property.” [17 U.S.C. §115(d)(11)(E)] Whether distributing unclaimed royalties to other copyright owners complies with the Due Process Clause of the Fifth Amendment, however, is a serious and open question.
The Collective “shall insure” that its records “are preserved and maintained in a secure and reliable manner” for at least seven years after creation or receipt. [17 U.S.C. §115(d)(3)(M)(i)] No more than once a year, a copyright owner may audit the records of the Collective to verify the accuracy of its royalty payments for any or all of the preceding three years (except that it may not audit the records for any calendar year more than once). [17 U.S.C. §115(d)(3)(L)(i)(I)]
Section 102(f) of the MWM Act directs the Copyright Office to prepare a report recommending best practices to “identify and locate musical work copyright owners with unclaimed accrued royalties,” “encourage musical work copyright owners to claim” those royalties, and “reduce the incidence of unclaimed royalties.” The report is due to Congress no later than 2 years after the initial designation of the Collective (which will occur no later than July 8, 2019).
The Digital Licensee Coordinator
The Act also authorizes the Register of Copyrights to designate a “digital licensee coordinator” to participate in certain proceedings. [17 U.S.C. §115(d)(5)] The digital licensee coordinator shall be a nonprofit, not owned by any other entity, endorsed and supported by “digital music providers and significant nonblanket licensees that together represent the greatest percentage of the licensee market for uses of musical works in covered activities … over the preceding 3 full calendar years” (again, “preceding” what is not specified; the best answer is preceding the designation date), and it must demonstrate it has or will have the administrative capability to perform its required functions. [17 U.S.C. §115(d)(5)(A)] As with the Collective, the Register must make the initial designation no later than July 8, 2019, and must revisit the designation every five years after that. [17 U.S.C. §115(d)(5)(B)(i),(ii)] Unlike the Mechanical Licensing Collective, however, “[i]f the Register of Copyrights is unable to identify an entity that fulfills each of the qualifications … the Register may decline to designate a digital licensee coordinator.” [17 U.S.C. §115(d)(5)(B)(iii)] In that case, statutory references to the digital licensee coordinator may be ignored; except that sometimes the statute allows digital music providers and significant nonblanket licensees representing more than half the licensee market to participate instead of the digital licensee coordinator. [e.g., 17 U.S.C. §115(d)(7)(D)(v)]
The digital licensee coordinator is authorized to perform the following eight functions: (I) create a governance structure, criteria for membership, and establish dues; (II) help enforce notice and payment obligations for the administrative assessment; (III) initiate and participate in proceedings before the Copyright Royalty Board with regard to the administrative assessment; (IV) initiate and participate in proceedings before the Copyright Office; (V) gather and provide documentation for use in proceedings before the Copyright Royalty Board; (VI) maintain records of its activities; (VII) assist in publicizing the ability of copyright owners to claim royalties for unmatched musical works; and (VIII) any other activities that are “necessary or appropriate” to fulfill its responsibilities. [17 U.S.C. §115(d)(5)(C)(i)]
Significant Nonblanket Licensees
A “significant nonblanket licensee” is defined as any entity that engages in covered activities outside the blanket license, if it exceeds a certain size: either (I) it makes more than 5,000 different sound recordings available on a single day, or (II) it has revenue that exceeds $50,000 in any calendar month, or $500,000 in any 12-month period. 17 U.S.C. §(e)(31). Each such licensee must identify itself to the Collective, 45 days after January 1, 2021 (or 45 days after the end of the first month in which it qualifies); and each must report its usage activity to the Collective and pay its administrative assessment each month, due 45 days after the end of the month. [17 U.S.C. §115(d)(6)(A)(i),(ii)] If an entity ceases to qualify, it may notify the Collective and cease reporting; but it must resume reporting again if it reaches the threshold size. [17 U.S.C. §115(d)(6)(A)(iii)]
If the Collective becomes aware of a significant nonblanket licensee that is not complying with the reporting or payment requirement, it must report that entity to the digital licensee coordinator [17 U.S.C. §115(d)(6)(B)]; and either may commence an action in U.S. District Court to obtain an injunction and damages. [17 U.S.C. §115(d)(6)(C)] Absent excusable neglect, the court must award treble the administrative assessment due, plus reasonable attorney’s fees and costs. [17 U.S.C. §115(d)(6)(C)(i)]
Subsection (d)(9)(A) provides that a blanket license shall supersede existing compulsory licenses for digital phonorecord delivery, with one exception. “On the license availability date [January 1, 2021], a blanket license shall … be automatically substituted for and supersede any existing compulsory license previously obtained under this section … to engage in … covered activities with respect to a musical work.” [17 U.S.C. §115(d)(9)(A)] (Recall that “covered activity” is limited to “digital phonorecord delivery of a musical work, including in the form of a permanent download, limited download, or interactive stream.” [17 U.S.C. §(e)(7)]) The exception is “that such substitution shall not apply to any authority obtained from a record company pursuant to a compulsory license to make and distribute permanent downloads unless and until such record company terminates such authority in writing to take effect at the end of a monthly reporting period, with a copy to the mechanical licensing collective.” [17 U.S.C. §115(d)(9)(A)]
This section is ambiguous, with two possible meanings. It could mean that if a digital music provider wants to engage in covered activity, it must either do so on an individual work-by-work basis, or obtain a blanket license; and that if it obtains a blanket license, that blanket license supersedes any individual licenses (except a permanent download license, which will continue in effect unless and until the record company terminates it in writing, in which case the blanket license will supersede it at the beginning of the next month). Or, it could mean that the substitution of “a blanket license” is mandatory, whether or not the digital music provider wants a blanket license or attempted to obtain one. Given that Congress used the words “may obtain a blanket license” in subsection (d)(2), and provided a procedure for obtaining the blanket license as an alternative to an individual license, the better reading is the first one; but litigation about the meaning of this section certainly is possible.
Subsection (d)(9)(B) provides: “Except to the extent provided in subparagraph (A), on and after the license availability date [January 1, 2021], licenses other than individual download licenses obtained under this section for covered activities prior to the license availability date shall no longer continue in effect.” Grammatically, there are four possible ways to read this section:
“[Except as provided in (A),] licenses // other than individual download licenses obtained under this section for covered activities prior to the license availability date // shall no longer continue in effect.” This is the only reading that definitely can be excluded, because it does not specify which licenses “shall no longer continue in effect.” Surely Congress did not intend to nullify all licenses (or all those relating to musical works).
“[Except as provided in (A),] licenses // other than individual download licenses obtained under this section for covered activities // prior to the license availability date shall no longer continue in effect.” In other words, unless subsection (A) applies, one needs to get a new license [individual or blanket] after January 1, 2021. This reading is possible, but unlikely, because there is an implicit verb missing. If this reading was intended, it should have said “licenses … made or obtained or entered into prior to the license availability date shall no longer continue in effect.”
“[Except as provided in (A),] licenses // other than individual download licenses obtained under this section // for covered activities prior to the license availability date shall no longer continue in effect.” This further limits the licenses that “shall no longer continue in effect” to those “for covered activities” (meaning that licenses for things other than digital phonorecord delivery would continue in effect); but it suffers from the same grammatical problem as the second reading, thereby also making it unlikely.
“[Except as provided in (A),] licenses // other than individual download licenses // obtained under this section for covered activities prior to the license availability date shall no longer continue in effect.” This reading makes the most sense grammatically, as the clause “prior to” now modifies the verb “obtained.” Under this reading, section 115 licenses for covered activities (digital phonorecord delivery) obtained prior to January 1, 2021 shall no longer continue in effect, with two exceptions: 1) individual download licenses may continue in effect; and 2) licenses that are automatically substituted under (A) or that are exempt from automatic substitution under (A) may continue in effect. The only significant objection to this reading is that it is redundant, because the “individual download licenses” that continue in effect under subparagraph (B) overlap substantially with the “permanent downloads” that may continue in effect under subsection (A). Nonetheless, this appears to be the best reading of this poorly-written subsection.
Subsection (d)(9)(C) provides: “A voluntary license for a covered activity in effect on [January 1, 2021] will remain in effect unless and until the voluntary license expires according to [its] terms … or the parties agree to amend or terminate [it].”
Subsection (d)(9)(D) provides that as of the enactment date (October 11, 2018), “the Copyright Office shall no longer accept notices of intention with respect to covered activities,” and that “notices of intention filed before [October 11, 2018] will no longer be effective or provide license authority with respect to covered activities.” However, if a valid notice of intention was filed with the Copyright Office before October 11, 2018, the person who filed it cannot be held liable for infringement for covered activities before January 1, 2021.
Limitation on Liability
If a digital music provider complies with certain requirements, then the copyright owner’s “sole and exclusive remedy” against the provider for engaging in covered activities between January 1, 2018 and January 1, 2021 is limited to the royalty for digital phonorecord deliveries established by the Copyright Royalty Judges. [17 U.S.C. § 115(d)(10)(A)] The four requirements are (i) within 30 days after first making a musical work available (or within 30 days after enactment), the provider “shall engage in good-faith, commercially reasonable efforts to identify and locate each copyright owner of such musical work,” including obtaining information from the sound recording copyright owner, and using a “bulk electronic matching process” provided by a third-party vendor (or its own such process, if comparable or better); (ii) the matching efforts must be repeated once a month “for so long as the copyright owner remains unidentified or has not been located”; (iii) if the matching effort is successful within the first 30 days, the provider “shall provide statements of account and pay royalties” to the copyright owner; and (iv) if the matching effort is not successful within the first 30 days, the provider shall accrue and hold royalties until they “can be paid to the copyright owner or are required to be transferred to the” Collective. [17 U.S.C. § 115(d)(10)(B)] Accrued royalties (with statements) must be paid within 45 days after the end of the calendar month in which the musical work copyright owner was identified and located, with monthly statements and royalties thereafter. [17 U.S.C. § 115(d)(10)(B)(iv)(II)] If the owner is not identified or located by January 1, 2021, then the provider must transfer the accrued royalties (with statements) to the Collective no later than February 15, 2021. [17 U.S.C. § 115(d)(10)(B)(iv)(III)] No late fees accrue until royalties are due to be paid or transferred. [17 U.S.C. § 115(d)(10)(B)(v)] The statute of limitations for these royalties is the later of three years from the date of accrual, or two years from January 1, 2021. [17 U.S.C. § 115(d)(10)(C)]
The MWM Act extends the existing antitrust exemption to “negotiations and agreements between and among copyright owners and persons entitled to obtain a compulsory license for covered activities,” including with respect to the administrative assessment. [17 U.S.C. § 115(d)(11)(A)] Except for the administrative assessment and interim rates and terms, “neither the [Collective] nor the digital licensee coordinator shall serve as a common agent with respect to the establishment of royalty rates or terms.” [17 U.S.C. § 115(d)(11)(B)] The antitrust exemption also extends to the Collective’s administration of voluntary licenses, except that each copyright owner must establish its own rates and terms for voluntary licenses, without acting in concert with other copyright owners; each digital music provider must agree to rates and terms individually, without acting in concert with any other digital music providers; and the Collective must keep the voluntary licenses confidential (to avoid conscious parallel behavior). [17 U.S.C. § 115(d)(11)(C)] Finally, the Register of Copyrights may “adopt such regulations as may be necessary or appropriate to effectuate the” Act, subject to judicial review under the Administrative Procedure Act. [17 U.S.C. § 115(d)(12)(A),(B)]
Performing Rights Organizations
Performing rights organizations (PROs) license the right to publicly perform any of the musical works in their repertoire, collect the royalties, and distribute them to musical work copyright owners. The three longstanding PROs are ASCAP (American Society of Composers, Authors and Publishers, BMI (Broadcast Music, Inc.), and SESAC (originally the Society of European..