On Tuesday, Meta announced that it was joining the ranks of X and other counterparts, like YouTube, by scrapping its third-party fact-checking program and moving to Community Notes—but the substance of other significant updates in Mark Zuckerberg’s statement could mean far reaching implications for the state of antisemitism on Facebook, Instagram and Threads. Meta’s announced changes may mean that more Jew-hatred will be left online for longer and for a wider audience than ever before.
More onus on users to moderate, with no additional resources for responding to reports
According to Zuckerberg’s statement, Meta will be relying more on user reports to flag ‘lower severity violations’ of policy, rather than actively scanning and pre-emptively flagging such content with their own resources – perceivably, this includes hate speech and antisemitism. The Jewish community is currently experiencing one of the worst onslaughts of popularized Jew-hatred since World War II and the echo chambers created by engagement algorithms indicate that this anti-Jewish trend will not go out of fashion any time soon.
The American Jewish Committee’s 2023 survey found that only one in three Jewish people who experience antisemitism reported it to the relevant social media platform. That’s likely because most of the content that users flag to platforms is filtered through AI – which has not been expertly trained on recognizing Jew-hatred. Most users give up reporting content after being gaslit by a platform’s AI telling them that blatant hate speech does not violate the platform’s policy. If the platforms want the users to enforce their rules for them, where is the guarantee of increased human resources or better-trained AI to handle the influx of reports?
Relying more on user-based reporting for a minority community bodes a guarantee of more hatred against that community being left online. The lack of commitment to increase the likelihood that a user report will be viewed by a human moderator or with more accurate AI, coupled with a parallel change to increase ‘Civic Content’ (socio-political and political content) in the feeds of Meta users, means that hate speech, including against the Jewish community, will likely reach more viewers than before. What happened on Twitter after it was acquired by Elon Musk yielded exactly this.
The exceptions to this projected reduction in enforcement include antisemitic content that also meets other criteria, including incitement to violence policies (of which CyberWell tracked a dramatic increase on social media in 2024) and anti-terror policies, both of which Meta confirmed it intends to keep enforcing at scale with its automations and filters.
Additionally, the changes in the fact-checking program and treatment of misinformation and disinformation may mean that fewer Jewish, Israeli and pro-Israel accounts find themselves in an erroneous “Instagram jail” for sharing content about the rise in antisemitism, how it is affecting them, and the current multi-front war in Israel. For example, recently CyberWell helped an account that was penalized for sharing official CIA statistics showing the population of Gaza increasing by 2% in 2024 after the account was apparently sanctioned and restricted for sharing ‘misinformation’.
But as the abuse of Trust & Safety channels, including the mass reporting of accounts by cells of antisemites, detractors of Israel and their foreign-funded bot farms to silence Jewish voices has become part of the antisemitic climate online, these changes mean even less systematic consequences for those who target our community in digital spaces through DMs and comments.
In the final segment of his statement, Zuckerberg touts his new partner in the fight against censorship: the U.S. government. The digital space, which includes everything from social media recommendation algorithms to the ever-present and very powerful generative AI, is unregulated with little to no transparency requirements which leaves a vacuum for intervention, pressure and even censorship from the incoming Trump Administration the same way it did from the exiting Biden Administration. Social media platforms see an opportunity to systematically lower the bar on their Trust & Safety promises and reduce allocated resources to protect their users, and they are taking it—but that will have a big cost for the safety of the Jewish community at an already difficult and dangerous time.
Antisemitism has historically been some of the best performing ‘content’ of all time, from the Catholic Church’s blood libel in 15th century Italy, to Der Sturmer in the Third Reich and now through supercharged engagement algorithms across our favorite apps and websites. Regardless of your political affiliation, it is imperative that our community unite and engage around the crucial issue of social media reform and transparency requirements—now more than ever before. At CyberWell, we will continue to track and analyze Jew hatred online in real-time and build the tech-based solutions of tomorrow to support effective compliance with community guidelines and weed out AI bias at scale. The digital and real-world safety of the Jewish people depends on it.
Tal-Or Cohen Montemayor is the founder and executive director of CyberWell, an independent tech nonprofit working with social media platforms to monitor and catalog antisemitic rhetoric while improving enforcement and enhancement efforts through community standards and hate speech policies.
This op-ed is published in partnership with a coalition of organizations that fight antisemitism across the world. Read the previous article by Misha Keyvanfar.