Dear Chairwoman Maloney and Chairman Lynch:
Gab’s mission is to be the home of free speech online. It seeks to export American values and freedom to the maximum extent permitted by American law to Internet users around the world. Gab concurs with the Committee’s view that these values protect offensive and unpopular speech, but not illegal speech, such as threats. Unlawful speech is not and has never been allowed on Gab.
Gab’s position on this subject has been very consistent in numerous public statements dating back several years, including:
- A blog post published on August 23, 2019, titled “Gab’s Policies Positions, and Procedures for Unlawful Content and Activity On Our Social Network,” in which I wrote:
“[Gab’s] terms of service absolutely ban all illegal content, including threats, as influenced by American speech laws as well as American legislation and court rulings on the subject. … Offensive speech and unpopular speech are protected by both our rules and by American law. Illegal speech, such as a threat, is absolutely not and has never been allowed on Gab. We have zero tolerance for lawbreakers on our platform.
- A blog post published on December 10, 2020, titled “Gab’s Statement On the Website Targeting U.S. Election Officials” in relation to a website, thought to be a foreign influence operation, which threatened U.S. election officials including Chris Krebs and others, in which I stated:
“At 8:38pm on Wednesday December 9th Gab received an email from Walden Macht & Haran LLP notifying us of a Gab account, @EnemiesOfThePeople, that was in breach of our Terms of Service.
“The Gab account was created just a few days ago and featured the personal addresses, photos, and more of election and government officials, which is against our terms of service. The account also made direct threats of violence towards these individuals, which is against the law as well as our terms of service.
“This type of content has no place on Gab and we have a longstanding history of zero tolerance for illegal behavior. At Gab we believe that free speech and open discussion are the best ways to solve problems and disagreements, not violence.
“Within minutes of receiving the email alerting us to the existence of this account we took immediate action by backing up the account information for law enforcement and then terminating the account from our service.
“We took it one step further by alerting the Gab community to this behavior and noted that our community members should report this type of illegal activity to our moderation team immediately if they come across it.
“At 9:34pm, less than an hour after being alerted to the existence of this account, our attorney replied to Walden Macht & Haran LLP to let them know that we took immediate action to terminate the account.”
- A blog post published on August 11, 2022, in which I wrote an open letter to FBI Director Christopher Wray that
“Use of our platform for controversial but lawful speech is our raison d’etre. Using our platform to threaten anyone, including federal law enforcement, is totally unacceptable. When we are notified by law enforcement of potential threats to life and property, we take quick action… [a]lthough we never publicly comment on specific requests made of us by U.S. or international law enforcement– a policy we have to ensure law enforcement is comfortable communicating with us on public safety matters–the FBI knows how to reach us.”
Your letter states that you are “concerned that reckless statements by the former President and Republican Members of Congress have unleashed a flood of violent threats on social media that have already led to at least one death and pose a danger to law enforcement officers across the United States.”
Gab is responsive to law enforcement’s requests and sympathetic to law enforcement’s plight. For years, Gab’s staff have frequently found themselves on the receiving end of threats of death, physical injury, and other unpleasantness by bad actors across the political spectrum. Only a few weeks ago, following the public disclosure, in an arrest warrant, that Gab rendered disclosure of user data to the FBI in response to an emergency data request, we were threatened by anonymous person or persons who tried to blackmail us into ceasing our entirely reasonable level of cooperation with the FBI in regards to unlawful threats of violence posted on our platform. We believe very strongly that we have a role to play in reducing the volume of unlawful threats on the Internet by aggressively policing our platform for this type of illegal activity, while at the same time providing the most free and open space for debate and dissent on the entire Web.
Dealing with threats of violence is, unfortunately, standard operating procedure for any company hosting user-generated content at scale. All social media companies work hard to square their missions of connecting human beings with the vagaries and vicissitudes of human nature. Gab is no different, and shares the Committee’s concerns regarding violent rhetoric on the Internet.
Responsibility for dehumanizing rhetoric on the Internet starts at the top. Gab denounces the trend in recent years where political leaders of both parties – presidents, governors, and congressmen alike – condemn the other side’s supporters as enemies of the state. That there are vast millions of Americans who disagree with any given point of view, often vociferously, is normal and to be expected in a free and democratic society.
There are very few questions on which reasonable people cannot reasonably disagree. Even in relation the most extreme honestly-held opinions, the chasm separating a philosophical opponent from a public enemy is – or at least it should be – wide and deep. America’s leaders, of both parties, egged on by their most extreme supporters, appear to have lost the ability to tell the difference or the desire to try. As the temperature rises Gab will do everything in its power to ensure it behaves as a responsible corporate citizen and renders prompt assistance to law enforcement on public safety matters while remaining true to its corporate mission.
Jesus Christ is King
RESPONSES OF GAB AI INC TO THE INQUIRY FROM THE HOUSE COMMITTEE ON OVERSIGHT AND REFORM DATE 19th AUGUST 2022
Since August 8th, 2022, how many threats against federal law enforcement has your company identified on its platform?
Between August 8th, 2022 and September 2nd, 2022, Gab received 940 threat reports from its moderation system. Gab’s moderation team removed 173 posts containing unlawful threats and suspended 15 accounts.
While Gab is aware of a number of threats against federal law enforcement, and has taken swift action against threats against federal law enforcement that have been brought to its attention, the Company does not collect data on the identity of the target.
How many of these threats were removed from your platform? How many were reported to law enforcement?
Regarding the number of threats removed, please see the Company’s response to (1) above.
Regarding reports to law enforcement, while Gab can confirm that it has been in contact with law enforcement in the month of August regarding threats against law enforcement (as this is a matter of public record from a public arrest warrant), as a matter of Company policy Gab does not discuss nonpublic communications with law enforcement, even where it might be convenient from a public relations perspective to do so.
Gab has this policy for two reasons: first, due to the possible existence of nondisclosure orders; and second, because Gab wants law enforcement to know that Gab will handle confidential information with the utmost discretion in situations where nondisclosure orders cannot be obtained, for example in situations requiring emergency disclosure.
If you would like information on the Company’s interactions with law enforcement in the month of August, 2022 or at any other time, the Company refers you to the Department of Justice and the FBI.
How many of the threats reported to law enforcement were related to the search warrant executed on former President Trump’s Mar-a-Lago residence, to the FBI, or to the Department of Homeland Security?
Gab does not disclose its communications with law enforcement. Please see (1a) above. Additionally, this request requires speculation by Gab as to facts it cannot possibly know. It is impossible for Gab to know whether an individual publishing violent antigovernment sentiments was motivated by particular law enforcement actions or whether such an individual is simply expressing beliefs which were held prior to August 8th, 2022.
What are your company’s policies, procedures, and practices for removing threats from your platform and reporting threats to law enforcement?
Gab’s policy is contained entirely in its Terms of Service, found at http://gab.com/about/tos , which states in relevant part that:
You may use the Website only for lawful purposes and in accordance with these Terms of Service. You agree not to use the Website… [i]n any way that would violate any applicable federal, state, or local law of the United States of America (including, without limitation, any laws regarding the export of data or software to and from the US or other countries) and is not protected by the First Amendment to the U.S. Constitution (the “First Amendment”).
As a general rule, written expression that is protected political, religious, symbolic, or commercial speech under the First Amendment of the U.S. Constitution will be allowed on the Website. User Contributions absolutely must in their entirety comply with all applicable federal, state, and local regulations in the United States.
Without limiting the generality of the foregoing, User Contributions must NOT:
- Be unlawful or be made in furtherance of any unlawful purpose. User Contributions must not aid, abet, assist, counsel, procure or solicit the commission of, nor constitute an attempt or part of a conspiracy to commit, any unlawful act. For avoidance of doubt, speech which is merely offensive or the expression of an offensive or controversial idea or opinion, as a general rule, will be in poor taste but will not be illegal in the United States.
- Unlawfully threaten.
As to procedures, Gab is a small business with a close-knit team. Its procedures for dealing with this content are informal and not constituted in a policy document. Describing the process, user reports prompt reviews by a moderation team, and the moderation team exercises its discretion as to what sanction, if any, to impose.
Gab believes it can responsibly follow this commonsense approach to content moderation because, unlike its competitors, Gab has always sought to have simple-to-understand Content Standards which can be understood by, and implemented by, ordinary people applying ordinary powers of reason. Where a threat is concerned, most adults innately understand where the line is. In implementing this policy the Company has not required, and has not prepared, extensive policy manuals or training documentation for content moderators and the Company’s staff.
What is the threshold for a threat to be removed from your platform, and how does your company determine this threshold?
The Company relies on moderators to exercise their reasonable discretion in applying the Terms of Service. Much as in real life enforcement of the law, the Company grants broad discretion to moderators to make judgment calls to decide when they believe a violation of the Terms of Service warrants a response, and the level of the response. Repeat offenders or accounts which generate a high volume of bona fide user reports are more likely to get the attention of the moderators than others. In some cases, bad actors known to the Company to be terrorist actors – in particular the neo-Nazi terrorist groups Feuerkrieg Division and Atomwaffen Division – are known by Company staff to have unlawful purposes and are liable to be banned on sight.
What resources does your company use in identifying threats on your platform, including but not limited to the number of employees assigned to content moderation?
The Company’s main content moderation resource is its user community. Gab strongly encourages its user community to self-police. Its usual practice for content moderation is to rely on user reports to bring threatening content to its attention. The Company assumes, but is unable to confirm, that law enforcement data requests it receives also originate with citizen reports to e.g. the FBI’s tip line or local police. The Company has been clear in its public statements that it will respond promptly to any lawful demand served on it by U.S. law enforcement.
Cognizant that Gab’s competitors may utilize automated tools to prevent threatening speech from ever being uttered to begin with, Gab considers these ineffective for the following reasons:
-Gab’s mission is to promote freedom of expression on the Internet. Blanket, automated bans on certain words and phrases risk chilling lawful speech.
-Whether content containing threatening language is in fact an unlawful threat is context dependent and difficult for a third party such as Gab, which merely hosts content and does not ordinarily have insight into the states of mind of its users, to parse by itself. Automated tools also run the risk of suppressing lawful speech which appears to be unlawful on its face but is may in fact be arguably lawful on closer examination. See e.g. Watts v. United States, 394 U.S. 705 (1969), and for a more contemporary example Paul Chambers v. Director of Public Prosecutions,  EWHC 2157 (QB) (a.k.a. the “Twitter Joke Trial”).
-Gab does not agree with the proposition that lawful but inflammatory speech should be censored. Even an unlawful threat which falls short of direct incitement (see, e.g. Brandenburg v. Ohio, 395 U.S. 444 (1969)) is, quite simply, incapable of being the proximate cause of physical harm to anyone. The childhood adage that “sticks and stones can break my bones, but words can never hurt me” is instructive. Gab further believes that allowance for inflammatory speech provides valuable advance warning of possible unlawful activity which law enforcement can follow up for further investigation. It therefore believes that automated content moderation which prevents harmful speech from being expressed in the first place would be detrimental not just to human freedom, but also detrimental to public safety.
Has your platform experienced an increase in the number of threats related to law enforcement or a change in the nature or specificity of these threats since August 8th, 2022? If so, please explain.
See response to (1).
You also requested that Gab produce the following documents by September 2nd, 2022:
All documents regarding your Company’s policies, procedures, and practices in addressing threats and incitement of violence on your platform, including, but not limited to, training materials.
See Gab’s Terms of Service, located at https://gab.com/about/tos , and Gab’s response at (2) above. Gab relies on moderators to use common sense in applying its Terms of Service, and the Terms of Service have been written in such a way as to have an objective meaning and be understandable by ordinary people. This differs from the complex and subjective approaches to censorship of so-called “harmful but lawful” content adopted by Gab’s competitors such as Facebook and Twitter.
All advertisements shown alongside posts that were reported to law enforcement or removed by your company for threatening law enforcement.
This information is not logged. Accordingly no records exist.