In the days following the 2016 United States presidential election, many were left wondering how the country had become so divided. Never before had the voters on either side of the aisle come to the polls with not only different opinions, but different facts upon which those opinions where based. This realization led to the ongoing period of reflection that still envelopes the country. News sources have not been impervious to such reflection, and have begun to look into the vetting process they currently employ with respect to the “news” they publish. Aside from obvious ethical and professional standards, the 2016 presidential election provided perhaps the most jarring example of the effect of inadequately vetted, partisan news.
One such news source that has received scrutiny is Facebook, Inc. The social media giant had to reconcile the fact that it had become a primary source of news for millions of individuals, a role for which it was decidedly ill-equipped. As a potential solution, Facebook launched a “global fact-checking initiative” in December 2016. This initiative involves, in part, employing groups of fact-checkers to review news published on the site. When an article has been deemed to be uncorroborated or misleading, the fact-checkers are tasked with publishing an explanatory article, notifying the user that posted the misleading article and ensuring the misleading article is then shown less prominently on the site.
Facebook currently has 43 fact-checking organizations across the world, covering news in 24 different languages. However, the fact-checkers themselves are uncertain as to whether they are having a material impact. Facebook requires fact-checkers to sign non-disclosure agreements, but this has not stopped many from anonymously speaking up about Facebook’s lackluster procedure with respect to the fact-checking process. Editors have reported feeling underutilized, and admitted that fact-checking, despite outward appearances, is not a priority for the Facebook brass. In fact, editors have noted that certain fact-checking groups cease operations when nearing the payment cap, which is a cap on the number of fact-checks for which Facebook has agreed to pay in a given month. This cap on explanatory articles results in a backlog from month-to-month, and the current cap is not nearly enough to provide a thorough fact-check of many of the articles posted to Facebook each month. Indeed, one fact-checker noted that its firm had nearly 500 articles in queue to be checked at the end of a certain month.
The current initiative will have to undergo an overhaul, especially when the number of articles to be reviewed are combined with articles posted to Facebook’s other networks, such as WhatsApp and Instagram. Facebook is acutely aware of the shortcomings of the current process, but as Mark Zuckerberg and other executives begin to explore alternatives, it appears a solution to this Herculean task remains far off.
The United States Court of Appeals for the Sixth Circuit has ruled that a tweet posted by Hollywood actor James Woods contained sufficient ambiguity to avoid defamation liability to a plaintiff who claimed that Woods wrongly portrayed her as giving a Nazi salute at a Donald Trump 2016 presidential rally. The court held that the inclusion of a question mark could deem the tweet to be posing a question rather than expressing a statement of fact.
Portia Boulger, who supported Democratic presidential candidate Bernie Sanders in the 2016 election season, sued Woods for defamation after the actor tweeted a photo of Boulger alongside another photo of a woman making a Nazi salute at a Trump rally. Woods’ tweet implied that Boulger may have been a plant at the rally by tweeting “So-called #Trump ‘Nazi’ is a #BernieSanders agitator/operative?” It was later confirmed that the woman giving the salute was not Boulger, but another person. Woods deleted the tweet with an apology to Boulger. Boulger sued Woods for defamation, which an Ohio federal district court dismissed at the pleadings stage. Boulger appealed to the Sixth Circuit.
The Sixth Circuit analyzed the tweet under Ohio law’s “four-prong, totality-of-the-circumstances test” to determine whether Woods published a false statement of fact, and addressed “the issue of whether questions can (or cannot) be defamatory.” The four factors are: (1) the specific language used, (2) whether the statement is verifiable, (3) the general context of the statement and (4) the broader context in which the statement appeared.
Regarding the first factor, the appellate court found that some readers of the tweet likely viewed it as an insinuation that Boulger was the women in the photo giving the salute, but it seems equally plausible that the tweet was posing a question such that its content had a precise meaning. Analyzing the second factor, the court held that the tweet did not present an obvious example of a question that could be factually verified, which requires reviewing its context under the third and fourth factors. Woods is a tweeter of politically charged content with sarcasm, exaggeration and hyperbole—characteristics more likely to be seen in an opinion, rather than a statement of fact. Woods’ tweet with Boulger’s photo is reasonably susceptible to both a defamatory meaning—that Woods was asserting she was the woman giving the salute, and an innocent meaning—that Woods was merely asking his readers a question. Because the tweet could reasonably be read to have an innocent meaning, it is not actionable as a matter of law.
A concurring opinion, agreeing with the result, stated that application of the four-part test is awkward in this social media context. The analysis simply requires asking whether a reasonable reader would interpret the tweet as a genuine question. Regarding Woods’ tweet, the inclusion of the question mark assumes that the writer is asking a question, such that a reasonable reader would not interpret it as an implied statement of fact.
Last Tuesday, February 20, Twitter was sued for allegedly censoring speech by an “alt-right” user.
According to the complaint, plaintiff Jared Taylor is the founder of New Century Foundation, a 501(c)(3) that’s purpose is to “disseminate facts about race and race relations” and to “study the effect that immigration is likely to have on the changing demographic character of the [United States].” Mr. Taylor and his organization had been active Twitter users since joining in 2011. At one time, Mr. Taylor’s personal account had over 40,000 followers and his publication, The New Republic, had over 32,000.
On December 18, 2017, Twitter permanently suspended both of these accounts because, according to an e-mail Mr. Taylor received from Twitter, he was “violating Twitter’s Terms of Service, specifically the Twitter Rules against being affiliated with a violent extremist group.”
Mr. Taylor denies he has ever practiced, promoted or advocated violence against anyone or affiliated with any organization that does. To the contrary, he contends that, “[w]hile the Plaintiffs hold admittedly controversial positions, they have always expressed them—both on and off Twitter—in a lawful, civil and respectful manner.”
Mr. Taylor alleges that—contrary to Twitter’s justification, which he characterizes as pretext—his account was suspended “due to nothing more than the controversial nature of the speaker’s viewpoint, political beliefs and perceived political affiliations.” This presents a legal question that courts around the country are just beginning to address—assuming Mr. Taylor does not advocate violence or associate with groups that advocate violence, can Twitter suspend his account solely because it dislikes or disagrees with what he says?
The answer is not as straight forward as a casual commentator might believe. It is a common misconception that the First Amendment grants every American citizen the right to say whatever he or she wants on the internet. In truth, the First Amendment does not grant any rights; it simply restricts the government from “abridging the freedom of speech.” In general, the First Amendment does not prohibit private entities, like Twitter, from restricting or even prohibiting speech on private property.
Decades ago, California courts created an exception to that general rule, holding that the California Constitution goes further than the United States Constitution and protects speech in privately owned shopping centers. Private owners of so-called “public forums” can place reasonable time, place and manner restrictions on the speech that takes place on their property; but these private owners cannot restrict speech based on the content or viewpoints being expressed.
In his lawsuit, Mr. Taylor argues that Twitter is the “paradigmatic example” of a privately owned space that functions as a public forum. “Because Twitter is a protected public forum under California law,” Mr. Taylor argues, “Twitter may not selectively ban speakers from participating in its public forum based on disagreement with the speaker’s viewpoint, just as the government may not selectively ban speech that expresses a viewpoint it disagrees with.”
Twitter has not yet filed an answer to Mr. Taylor’s complaint. But, if California’s courts ultimately agree with Mr. Taylor, then Twitter and other social media companies (a substantial number of which are headquartered in California and therefore subject to the California Constitution) could face an onslaught of lawsuits by users who have been censored, disciplined or banned.
At the other end of the spectrum, national, state and local governments around the world are enacting laws that require social media companies to proactively search for and remove offensive speech from their platforms. Such laws are clearly in tension with lawsuits like Mr. Taylor’s. One can expect additional lawsuit as governments, social media companies and users try to strike the balance between unfettered speech and undue censorship.
Earlier this month, a Small Claims Court in Indiana ordered a defendant to pay $6,000 in damages for defaming a plaintiff on Facebook. According to an interview of the plaintiff’s attorney, published in the The Indiana Lawyer, “the court’s monetary award appears to set a precedent as the first reported judgment of its kind in the nation.”
The plaintiff in this allegedly groundbreaking case was Zerlie Charles. The defendant, Vickie Vest, dated the plaintiff’s son until his death in February, 2015. Shortly after his death, the son’s 2002 Chevrolet Silverado pickup truck was stolen. Police recovered the truck in a church parking lot.
On February 28, 2015, the defendant posted the following on her Facebook page:
Just have to say [a]ll the talk that's being said about Robert['s] things being stolen[,] [i]f it was stolen I don't know but I do know my truck was and [,] yes[,] Zerlie Charles had everything to do with it[,] that's facts [sic]. I didn't even get all my personal things out of the house before his mom went physco [sic]. Butt [sic] that's OK[.] I will be OK[.] I lost my soul mate [,] thrown out of his house[,] and had my truck stolen all in 2 weeks. So I really don't give a DAM! [sic] what Zerlie Charles has to say. I was there for Robert[.] [S]he had to have control[.] [W]ell she got it all now. And still ain't happy. Life goes on and will be great. She can talk all she wants and we all know she will because that how it is[.] I have our memories and a lot of wonderful ones that no one can take away! Not even Zerlie Charles!!!!!!!!!
The plaintiff claimed that, because of the defendant’s Facebook post, she could no longer sleep at night, her reputation had been “ruined” and some of her good friends did not “come around anymore.” So she filed a complaint against the defendant for defamation per se. And, last October, the Indiana Court of Appeals decided the case in the plaintiff’s favor.
Similar to the common law in many states, under Indiana law, a plaintiff asserting a claim of defamation must prove (a) the defendant made a defamatory statement, (b) that statement was published, and (c) the plaintiff suffered damages as a result. Defamation per se arises when the statement, without reference to extrinsic evidence, imputes criminal conduct; a loathsome disease; misconduct in a person’s trade, profession, office, or occupation or sexual misconduct. In those cases, damages are presumed. However, truth is a complete defense.
The Indiana Court of Appeals held that the defendant’s statement—“I do know my truck was [stolen] and[,] yes[,] Zerlie Charles had everything to do with it [,] that's facts [sic]”—was defamation per se because it (a) “quite clearly” imputed criminal conduct—(i.e., stealing a truck), (b) was published on Facebook (where at least eleven people read and “liked” it) and (c) was false.
On remand, the Small Claims Court awarded $6,000 in damages, the maximum allowed under the Small Claims Court’s jurisdictional limits.
In his interview with the The Indiana Lawyer, the plaintiff’s attorney issued what should by now be a familiar warning: “A lay person probably feels they have a First Amendment right to say whatever they want to, but that’s definitely not the case if you’re harming people with what you’re saying.”
On January 2, Paris Hilton announced, via Instagram, that she was engaged. Hilton quickly followed her announcement with several close-up photos of her $2 million engagement ring. In each photo, she tagged the jeweler that created the ring, Greene & Co.
Given the hefty price tag, many have speculated that Hilton and her fiancé bartered social media exposure in exchange for a discount. In a recent People magazine article, Michael Greene of Greene & Co. was quoted saying he would not disclose how the ring was paid for and specifically would not discuss whether there was a promotional deal involved.
More recently, Kanye West’s clothing company, Yeezy Apparel, launched a social media campaign featuring Instagram posts by celebrities and influencers wearing the Yeezy Season 6 collection. As the New York Times explained in a February 1 article, the photos were particularly attention-grabbing because they “[recreated pictures of Kim Kardashian modeling Yeezy Season 6 apparel], but with women like Paris Hilton, Sarah Snyder (best known as Jaden Smith’s ex-girlfriend) and Sami Miro (ex-Zac Efron) all dressed up as Kim-a-likes, complete with long platinum wigs.” Each post was accompanied by the caption “#YEEZYSEASON6.” The Times’ fashion critic lauded the campaign as “the most successful thing Mr. West has ever done in fashion.”
While many observers have dubbed these promotional schemes “brilliant,” there is one agency that may be unimpressed: the Fair Trade Commission (FTC). Last year, the FTC updated its guidance regarding social media endorsements and stepped up its enforcement efforts. The thrust of the FTC’s guidance is clear: if an endorser receives any material benefit in exchange for a post, that benefit should be clearly and conspicuously disclosed.
The FTC sent “educational” letters to more than 90 celebrities, athletes, influencers and marketers encouraging them to comply with the requirement of clear and conspicuous disclosures. Twenty-one of the educational letters were followed by “warning” letters, which cited specific posts the FTC believed were not in compliance. The FTC also initiated its first law enforcement action against an individual online influencer.
There are no special words that must be used. But the FTC has suggested that including “#sponsored,” “#promotion” or “#ad” is a good start. Simply saying “thank you, X company,” “#partner” or “#sp” is not enough.
None of the “brilliant” promotions described above included any such disclosure, which may land the brands and influencers in hot water. But it is relatively simple for you to avoid the same legal pitfall. If you receive any payment, product or discount in exchange for a social media post, disclose it. And, if your business is giving out money, free products or discounts in exchange for social media posts, then make sure those posts include an appropriate disclosure.
Yesterday, January 30, Facebook announced a new policy prohibiting “ads that promote financial products and services that are frequently associated with misleading or deceptive promotional practices.” The new policy expressly prohibits any and all advertising for binary options, initial coin offerings and cryptocurrency.
In the announcement, Facebook explained that the new policy “is part of an ongoing effort to improve the integrity and security of [Facebook] ads, and to make it harder for scammers to profit from a presence on Facebook.”
Facebook’s Advertising Policies already prohibit “deceptive, false, or misleading content, including deceptive claims, offers, or methods.” Previously banned content includes:
“False or misleading claims about product attributes, quality, or functionality”;
“Setting confusing or misleading expectations for delivery times”;
“False or misleading claims about return processes”; and
“False or misleading claims about the services provided.”
For example, the phrase “Earn profits every 2 weeks” is allowed; “Earn 15% profits every 2 weeks” is not.
The newly announced prohibition goes much further, barring any advertising by binary options, initial coin offerings and cryptocurrencies, regardless of whether a specific advertisement is actually deceptive, false or misleading. According to Facebook, the new policy is “intentionally broad.” Facebook claims a categorical ban (rather than one that assesses the merits of cryptocurrency advertisements on a case-by-case basis) is necessary because “many companies who are advertising binary options, ICOs and cryptocurrencies . . . are not currently operating in good faith.”
Facebook has promised to revisit this issue as its capabilities for detecting and removing deceptive advertisements improve. In the meantime, however, all advertising of these types of financial products and services is prohibited.
The announcement was accompanied by a warning that Facebook is ramping up enforcement of its policy against deceptive, false and misleading content across all of its platforms, including Facebook, Audience Network and Instagram, and is actively developing better methods for detecting and removing such advertising from those platforms.
On Saturday, January 27, the New York Times published an exposé of Devumi, “an obscure American company” that “has collected millions of dollars in a shadowy global market for social media fraud.” According to the Times, Devumi has “an estimated stock of at least 3.5 million automated accounts, each sold many times over.” Here’s the big problem: 55,000 of those accounts “use the names, profile pictures, hometowns and other personal details of real Twitter users, including minors.”
Within hours of the article’s publication, New York State Attorney General Eric T. Schneiderman opened an investigation into Devumi. He tweeted that: “Impersonation and deception are illegal under New York law. We’re opening an investigation into Devumi and its apparent sale of bots using stolen identities.”
The Times article specifically named a number of prominent actors, models, athletes, reality television stars, journalists, politicians, and — yes — business executives who were customers of Devumi. It does not appear that any of Devumi’s customers are being investigated by the Attorney General at this time, but it is not hard to imagine that criminal and, more likely, civil claims may be forthcoming against companies that knowingly purchase social media followers that are impersonations of real people. Beyond legal liability, purchasers of social media followers risk deactivation of their social media accounts. Twitter’s Rules, for example, expressly prohibit the purchase of followers, retweets, and likes. At the very least, being exposed as a purchaser of social media followers is a public embarrassment that most companies would like to avoid.
How can your business protect itself going forward? Most of the Devumi customers who provided comments in response to the Times article claimed they had no knowledge that followers had been purchased; they claimed the transactions were authorized by a rogue employee, family member, or friend. So, one thing your company can do right now to protect itself is adopt written policies prohibiting employees and affiliates from buying social medial followers, likes, retweets/reposts, comments, or anything else that artificially inflates your social media presence.
If your organization outsources its marketing and public relationship functions, then your contracts should expressly prohibit third parties from purchasing social medial followers on your behalf. It should also include an indemnification provision that requires your marketing or public relations firm to defend and indemnify you against any investigation or claim related to the unauthorized purchase of social media followers.
We have posted pieces on several recent cases in which courts have addressed whether and how an anonymous blogger should be unmasked. Courts have reached conflicting results when balancing the alleged harms caused by anonymous posts against the speaker’s First Amendment rights. On November 28, 2017, the United States Court of Appeals for the Sixth Circuit became the first appellate court to weigh in on the issue. The Sixth Circuit addressed whether a plaintiff that prevailed in a copyright infringement lawsuit is entitled to injunctive relief that would include the unmasking of the John Doe defendant, who posted the company’s copyrighted materials on his blog. Signature Management Team LLC v. John Doe, No. 16-2188 (6th Cir. Nov. 28, 2017).
Signature Management Team LLC (“Team”) sells materials designed to help individuals profit in multi-level marketing materials. John Doe anonymous runs a blog that criticizes multi-level marketing companies. Doe posted a hyperlink to an edition of a book copyrighted by Team, which led Team to sue for infringement. Team sought judicial relief disclosing Doe’s identity, Doe’s destruction of all copies of the book in his possession and a permanent injunction barring Doe’s infringement use of the book. Doe responded by raising a fair use defense against the infringement claims and asserted a First Amendment right to speak anonymously. During discovery, Team moved to compel Doe’s identity. The trial court concluded that unmasking the anonymous speaker to Team could impact Doe’s defenses in the litigation, but it did order Doe to reveal his identity to the court and to Team’s lawyers, subject to a protective order preventing Team from learning Doe’s identity. When the case was reached on the merits, the trial court found for Team and had to determine the appropriate order. The trial court found that unmasking Doe was unnecessary because Doe represented that he would commit infringement again and had destroyed all copies of the book in his possession. Team appealed the trial court’s refusal to unmask Doe.
On appeal, the Sixth Circuit issued a split 2–1 ruling. Writing for the majority, Justice Helene M. White noted that “no case has considered the issue presented here—whether and under what circumstances a court can properly protect a party’s anonymity after judgment.” The fact that liability was established “is an important distinction. The prejudgment cases often deal with a plaintiff’s need to unmask a defendant to effectuate service of process . . . .” Regarding the issues before the court at this stage, Justice White wrote that the entry of a final judgment negates concerns that the unmasking could impair a defendant’s ability to defend itself in the litigation. Even so, there may not be a practical need for the post-judgment unmasking of an anonymous defendant who voluntarily complied with the relief to prevent further harm.
The majority ruled that the trial court applied too protective a standard in its ruling declining to unmask Doe. The trial court balanced factors developed in connection with pre-judgment proceedings. The majority stressed that the trial court failed to recognize that “very different considerations apply” after the entry of a final judgment on the merits, particularly the presumption in favor of open judicial proceedings. Nonetheless, the majority concluded that there are still factors suggesting that Doe may retain the right to remain anonymous, especially if an unmasking order would unmask him in connection with both protected and unprotected speech and might hinder his ability to engage in anonymous speech in the future. The Sixth Circuit remanded the case back to the trial court for reconsideration of the unmasking issue applying the concerns and factors identified in the majority’s opinion.
In a sharply worded and succinct dissent, Justice Richard F. Suhrheinrich criticized the majority for acting like “an overprotective parent.” The dissent stated that Doe should not be shielded from the consequences of his infringement actions, which are not protected speech under the First Amendment. Doe could have preserved his right to speak freely and anonymously by doing so without committing copyright infringement. The dissent contended that no balancing is necessary and that the proper course is to remand the case back to the trial court with an instruction to order the revealing of Doe’s identity.
We will monitor the proceedings on remand. This is not the last word in this case, and we expect to see similar issues continuing to arise in other cases with the proliferation of Internet speech.
On November 8, the United States Court of Appeals for the Ninth Circuit affirmed a lower court’s order compelling Glassdoor, Inc. to comply with a grand jury subpoena and disclose information about certain users who posted anonymous reviews on Glassdoor’s website. Glassdoor argued that complying with the subpoena would violate its users’ First Amendment Rights to associational privacy and anonymous speech. This is the latest ruling among other recent court decisions addressing the issues of anonymous speech on the Internet. United States v. Glassdoor, Inc., No. 17-16221 (9th Cir. 11/8/17).
Glassdoor operates Glassdoor.com, a website where employers promote their companies to potential employees, and employees post reviews of what it is like to work at the companies. Employees rate the employers in a variety of categories and describe workplace environments, salaries and interviewing practices. The reviews are anonymous, but users must provide Glassdoor with their e-mail addresses, though the addresses do not appear on the site.
An Arizona federal grand jury is investigating a governmental contractor that administers two Department of Veterans Affairs health care programs. As of March 2017, current and former employees of the subject company had posted 125 reviews on Glassdoor.com, including criticisms of its management and business practices. The government served Glassdoor with a subpoena that ordered it to provide the grand jury with “Company Reviews” and associated “reviewer information.” Glassdoor notified the government that it believed that the scope of the request implicated First Amendment concerns, and the government agreed to limit the request to reviewer information associated with eight exemplar reviews. Glassdoor still objected and sought that a trial court quash the subpoena. The trial court denied Glassdoor’s motion to quash, holding that Glassdoor had not shown that the grand jury investigation was being conducted in bad faith. Glassdoor appealed to the Ninth Circuit.
On appeal, the Ninth Circuit rejected Glassdoor’s argument that its users’ First Amendment associational rights were violated. Glassdoors’ users are necessarily strangers to one another because they are anonymous. They do not “discuss” employment conditions, but rather independently post their individual views.
Applying Supreme Court precedent to determine whether the subpoena violated the First Amendment, the Ninth Circuit held that Glassdoor did not show that the grand jury investigation is being conducted in bad faith or that the government’s subpoena is intended to harass. There is a substantial connection between the subject matter of the investigation and the identifying information of the eight users whose Glassdoor posts allude to potentially fraudulent behavior. The court concluded that any incidental infringement on Glassdoor’s users’ First Amendment rights is “no more drastic than necessary to vindicate compelling interests.”
Social media seems to have earned a bad rap lately. Oversharing has become the norm, cyberbullies are able to reach a wider audience and accounts can be hacked easily. However, finally, social media is getting recognition for its power to do good for the community in times of desperate need. Hurricane Harvey battered areas of Texas, leaving hundreds of residents stranded in waist-deep flooding. Unsure of when rescue efforts would be able to reach them, many victims turned to Twitter, Facebook and Instagram to request assistance. Given the flow of information in real time, some victims posted their address asking for rescue, either to emergency officials or to the general public. An online map titled “Harvey Relief” used geo-locational technology to allow victims to drop a pin on a location and request help. The map crowdsourced rescue requests through other platforms. With 911 fielding over 56,000 calls in just 15 hours, the posts enabled good Samaritans to assist rescue operations in aiding those stranded.
Social media also provided another source to disseminate needed information to the public. Houston’s Chief of Police Art Acevedo tweeted updates and videos from around the area. Harris County Sherriff Ed Gonzalez also used social media to alert the public to those in need, including a pregnant woman going into labor. People also used social media to let loved ones know they were safe.
Additionally, people were able to turn their homes into shelters through the vacation rental website, Airbnb. The site created a mechanism for people to turn their homes into shelters for free. The site also stated that it will waive fees for victims of Harvey between now and September 25.
Social media also became a great partner for raising funds to support the victims and the rescue efforts. For example, J.J. Watt of the Houston Texans has raised more than $4 million by posting videos of himself to Twitter. Other crowdfunding sites also created pages to accumulate donations for Harvey victims.
However, as much as social media has helped, it has also been marked by false information about rescue operations and incorrect photos of the disaster. Users should take caution in relying solely on information found on social media platforms and make sure to donate only to reputable organizations.
The last time a major hurricane made landfall in the United States, Twitter and many other social media sites did not even exist. Now, with Harvey continuing to rattle on the region, Twitter and other social media platforms are showing their true power to connect victims with loved ones, organize rescue efforts and raise funds for relief.