Amid Elon Musk’s ongoing dispute with the Anti-Defamation League (ADL), in which the ADL claims that X is now allowing more hateful content to be distributed in its app, while X, and Musk, claim that hate speech is well down on previous levels, X itself has published an official statement in order to underline its stance against antisemitism in any form.
As per X:
“Our teams at X constantly listen to feedback from users and maintain continuous dialogue with outside organizations to ensure our policies and enforcement balance free expression with platform safety. We firmly believe these two values can coexist, and we work tirelessly to achieve that goal. As part of this, we are committed to combating hatred, prejudice and intolerance – particularly when they are directed at marginalized and persecuted groups. That means taking proportionate action on content that violates our Rules.”
In support of this, X claims that it’s expanded its “Violent & Hateful Entities” policy, and refreshed its policies around violent speech “to reflect our zero-tolerance approach”. X also says that it continuously adds new slurs and harmful terms to its operational guidelines, to ensure that it’s covering “the evolving landscape and use of language to target members of protected categories, including members of the Jewish community”.
“Through our policies and enforcement, active training of agents, and working alongside our partners and users, X is committed to combating antisemitism on our platform. Our work is ongoing and we’ll continue to make investments in this area.”
The statement, it’s worth noting, was published late on Friday afternoon, which is generally the time that press releases go to die.
But that’s an aside, what really matters here is the substance, and as noted, amid ongoing concerns about X’s new “freedom of speech, not reach” approach, which has seen it update its platform rules, and lean more into de-amplification over removal, while also re-instating many previously banned users, X claims that organizations like the ADL are costing it billions in lost ad revenue, due to the false narrative that it’s continually reinforcing, which suggests that hate speech has actually increased in the app.
So, to clarify. Ever since Musk purchased Twitter, the ADL has been monitoring the platform’s various policy changes, amid concerns that Musk’s personal views on content moderation could lead to a rise in hateful content.
Back in May, the ADL published a review of Twitter/X since the Musk acquisition, and noted the following:
“[Our review] has found Twitter does not enforce its policies on antisemitism, even when flagged content openly incites violence, as we also found in our recent Holocaust Denial Report Card. And, of course, strong enforcement is only possible when teams responsible for trust and safety are adequately supported and resourced. Twitter, unfortunately, has eliminated most of its people responsible for content moderation, and it also disbanded its Trust and Safety Council, a volunteer group of independent civil society experts who advised the platform on curbing harms, of which ADL was a member.”
The ADL has also reported that X’s content policies are not adequate to combat antisemitism and hate speech, while it’s also found that QAnon-related content “has surged on Twitter since Elon Musk’s takeover of the platform”, among other reports.
X’s own data, meanwhile, suggests that hate speech instances are on the decline, due to an update in the way that it tracks such in the app.
Back in March, X partner Sprinklr outlined its own findings on instances of hate speech within X posts, in which it found that 86% of X posts that included slurs were not actually considered harmful or intended to cause harm.
“Sprinklr’s new toxicity model analyzes data and categorizes content as “toxic” if it is used to demean an individual, attack a protected category or dehumanize marginalized groups. Integrating factors such as reclaimed language and context allowed our model to eliminate false positives and negatives as well.”
Based on this, and using a list of 300 slur terms tracked by Sprinklr’s model, 86% of such mentions were actually, it claims, used in a non-offensive way.
Therefore, Sprinklr claims, any modeling that counts slur terms by pure mention volume is flawed, which it claims is why many reports have suggested hate speech is on the rise in the app, when it actually isn’t.
Which may be true, though 86% seems incredibly high. But nevertheless, that’s what X has run with, while it also reported back in July that “more than 99.99% of Tweet impressions are from healthy content, or content that does not violate our rules”.
So, according to X, only 0.01% of the posts you see on X are going to include any kind of hate speech or rule-breaking content.
Which would be a record-high performance rate on this front, and with X also cutting 80% of its staff, including, as the ADL notes, at least some of its moderation team, as you can image, there’s a level of skepticism around these claims, and whether X is actually, possibly meeting these levels of moderation.
The ADL says that it’s not, and it continues to produce evidence to support its stance, yet it is only using a small sample size within its analysis, which could still mean that X is achieving the above-noted numbers.
But X hasn’t published any actual data to support such, just the topline figures themselves, which is why they remain in dispute. X could avoid this, by publishing the full data, and allowing external analysis of its claims. But it hasn’t, and it likely won’t, though it may have to if Elon Musk moves to sue the ADL over lost ad revenue, as he’s more recently threatened.
It’ll be interesting to see what the data shows, if that actually happens, though my guess is that Elon will back away from legal action in this case, and will be hoping that he can bully the ADL into silence, while also showing potential ad partners that he’s confident in his counter-claims.
X’s ad revenue is down 60% in the U.S., though Musk has also stated that it’s improving in Asian markets, reducing its reliance on U.S. ad spend. X is also now generating more income from subscriptions and its higher-priced API access, and in combination, Musk claims that X is no longer as reliant on U.S. advertisers either way, even if they choose to withhold their ad spend.
But the vast majority of X’s ad income still comes from U.S. businesses. And clearly, given Musk’s angst against the ADL, it remains an important revenue stream.
The next step, then, would be for X to produce the full evidence to silence negative reports, and until that happens, I suspect that most advertisers will remain skeptical of X’s statements in this respect.
So rather than PR statements, X could just produce the data. It hasn’t yet.