TikTok has long attracted scrutiny from the United States government, with some officials deeming the Chinese-owned social media app a security risk. Increasingly, other influential groups are also criticizing the platform, claiming that it allows for a variety of harmful content to proliferate—especially content related to the current war between Israel and Hamas.
Fortune has learned that on November 16, about 40 mostly Jewish tech leaders met with TikTok CEO Shou Zi Chew and others on the company’s leadership team to express their concern over data they say shows that TikTok’s algorithm is favoring content that supports Palestine over pro-Israel content at a disproportionate rate. (Others, including celebrities and creators, have also met with TikTok to express similar worries in recent weeks.) The group, which includes the founders of Gusto, Bonobos, and Tinder, plus partners from firms like Techaviv and Bloomberg Beta, among other tech players, says that their goal was to get answers to what they see as an unexplainable discrepancy. They also wanted to share concerns over a rise in antisemitic incidents across the country and any possible correlation to online content distributed by social media platforms–particularly TikTok, which has become the de facto search engine for Gen Z users. Lastly, they were hoping to convince TikTok to take action too: To reexamine its algorithms and its policies.
“For every view of pro-Israel posts, there are about 54 views of pro-Palestine posts,” says Anthony Goldbloom, the founder of Sumble, a startup that curates public data for a variety of applications, and the mastermind of the November meeting with TikTok. “If TikTok was just a mirror reflecting back what people believe, it shouldn’t be a 54:1 ratio.”
In the weeks since the war broke out, Goldbloom, who sold his first data-centric startup, Kaggle, to Google in 2017, has taken it upon himself to scrape and parse TikTok content in the hopes of better understanding its reach and impact. His approach has been to look at engagement levels of posts with hashtags like #freePalestine versus #standwithIsrael, the results of which he says have left him puzzled.
“The volume of this content is so high on this platform, it just seemed crazy to me,” says Goldbloom.
Goldbloom had met Chew, TikTok’s CEO, at an event several months earlier. After digging into the data, he emailed Chew a slide showing his findings, and says there was some limited back and forth between the two. He eventually asked for a meeting, gathering other techies with similar concerns about TikTok for the call with the app's leadership team, which included TikTok's heads of operations and public policy, in addition to its CEO.
By all accounts, the call, which took place over Zoom, was largely cordial. (An exception: At one point, one of the participants told Chew, “I don’t believe I can trust you," according to someone who was at the meeting). But the group of techies were strong-worded in their concerns over the type and volume of war-related content prevalent on TikTok, and what impact it could have on stoking more antisemitism. The main gist: The tech executives wanted to understand how content on TikTok could lean so heavily pro-Palestinian (the group contends that even in Israel, the ratio of engagement with the top pro-Palestine hashtags to the top pro-Israel hashtags are 2:1). They also wanted to push TikTok to reexamine its community guidelines, arguing that even some posts that aren’t technically in violation of current rules could be leading to harm by spreading highly biased and spurious information that causes users to form antisemitic views or commit antisemitic acts.
TikTok says 'blunt comparisons' of hashtags is flawed
TikTok, for its part, has provided multiple answers and explanations to the above questions, both publicly and privately.
To begin with, the company takes issue with Goldbloom’s methodology, saying in a recent blog post that “blunt comparisons of hashtags is severely flawed and misrepresentative of the activity on TikTok.” Case in point: TikTok has said that while hashtags like #standwithIsrael may be associated with fewer videos than #freePalestine, the former has 68% more views per video in the US, which means more people are seeing the content.
In the same blog post, published on November 13, TikTok also vehemently denied that its algorithm is artificially skewed in any way. “The content people see on TikTok is generated by our community and recommendations are based on the content people have previously engaged with,” the company said. “TikTok does not ‘promote’ one side of an issue over another.”
Still, the social media app hasn’t tried to deny the overall imbalance of views for content that’s categorized under pro-Palestine versus pro-Israel hashtags. It simply asserts that this is not the product of any kind of intended or unintended bias in its algorithms, and that there are logical explanations for this.
A TikTok spokesperson who spoke to Fortune on background pointed to several reasons for the discrepancy of engagement levels. For starters, says the company, hashtags like #freepalestine existed long before the current war started—and before more recent pro-Israel hashtags were created—which means the body of content tagged with top pro-Palestine hashtags is much bigger than those on the pro-Israeli side. In addition, TikTok says that millions of users in the Middle East and Southeast Asia account for a significant percentage of views on hashtags, (the suggestion being that these regions skew pro-Palestinian as a population). Then there is the matter of the overall demographic of TikTok users, and their political leanings.
To be sure, TikTok’s users skew younger, and there is a proven generational divide when it comes to public opinion about the Israeli-Palestinian conflict, with multiple polls and surveys showing that younger people are more likely to sympathize with Palestinians versus Israelis when compared to older generations. But, says Goldbloom, these stats still don’t explain the extent of the divide in engagement with pro-Palestinian versus pro-Israeli content on TikTok which his data suggested: “Does this mean 98% of their U.S. users are pro-Palestinian?” he asks. (Indeed, no survey Fortune has seen would suggest that the divide is so steep, even among young Americans.) Others who joined the meeting were also left with more questions than answers as well, and a general sense of disappointment that while they were heard by TikTok’s higher-ups, the company doesn’t appear poised to act on any of their concerns.
An 'incredible diplomat' already in the hot seat over China
“The question is, do you want to look and say there’s a lot of reasons to explain this, or do you want to say that you’re troubled by the results and feel compelled to dig in to the type of information that’s [being] distributed and the impact that information appears to have,” says David Fischer, the former chief revenue officer of Meta, and one of the participants of the recent meeting.
TikTok says it has taken “continued, aggressive actions” to protect its community, adding Hebrew- and Arabic-speaking moderators and removing more than 1 million videos that promote misinformation, terrorism, and hate speech since October 7, and meeting with groups on both sides to hear them out in an attempt to improve platform safety.
"This is an extremely difficult time for millions of people around the world and in our TikTok community,” a company spokesperson told Fortune. “We feel it's important to meet with and listen to creators, human rights experts, civil society and other stakeholders to help guide our ongoing work to keep our global community safe."
Fischer says that TikTok’s executives were sympathetic to the group, and adds that he can understand “what it’s like to be on the other side” (referring to how difficult it is for a social platform to make changes). Another participant, who did not want their name mentioned, referred to Chew as an “incredible diplomat,” adding that they walked away from the 45-minute meeting with the conclusion that “we have to look to the U.S. government to act.”
This isn’t the first time that social media platforms have had to grapple with criticisms of the potential impact of the content they distribute during times of intense geopolitical strife. (An important note: Anti-Muslim views and acts are also on the rise, and some have criticized TikTok and other social platforms for unfairly censoring or limiting the reach of pro-Palestine content.) But in TikTok’s case, the scrutiny is greater: For the last few years, the company has found itself under the crosshairs of another geopolitical clash, between the U.S. and China. And while the group of tech executives who recently met with the company’s leadership team are pushing the company for more transparency and to revisit its content policies, some in the tech community were pushing for much more restrictive measures well before the current war in the Middle East broke out.
“The fundamental problem is that TikTok has two masters: the U.S. government and the Chinese Communist Party,” Jacob Helberg, senior policy advisor to the CEO at Palantir Technologies, and a long-time critic of TikTok, writes in an email. “And the laws on censorship and surveillance in China directly conflict with American laws protecting free speech and personal privacy. It is untenable for any company to simultaneously comply with two legal systems that have irreconcilable contradictions.”
TikTok has long said it does not share data or have connections with the Chinese Communist Party. But that hasn’t assuaged the scrutiny, which only appears to be growing under the current conflict.