While false and misleading election narratives are surging on social media in the lead up to the 2024 U.S. presidential election, major platforms have weakened their rules against disseminating election misinformation, according to a new ADL (Anti-Defamation League) report. Researchers at the ADL Center for Technology and Society (CTS) have also found that some platforms – primarily X (formerly Twitter) – have rolled back enforcement, including of stated exemptions for “high profile” accounts.

Ahead of the upcoming U.S. presidential election, a flood of narratives has surged that rehash antisemitic tropes and push anti-immigrant misinformation and ADL researchers evaluated whether the four key platforms where hateful election misinformation was found (X, Facebook, TikTok, and YouTube) implemented any of ADL’s previous recommendations or made other changes to their policies or enforcement.

“Once again, extremists and conspiracy theorists are leveraging social media platforms to spread false election narratives targeting marginalized groups, in particular Jews and immigrants,” said Jonathan Greenblatt, ADL CEO and National Director. “These narratives that include claims that Jews or Israel secretly control U.S. elections are not only false and grotesque but also dangerous. We call on social media platforms to act now to protect all communities and to ensure that hateful election misinformation is never tolerated.”

ADL found that these four platforms continue to make exceptions for some kinds of misinformation, such as false allegations of past voter fraud, and that their policies are generally insufficient to prevent coordinated disinformation campaigns, such as users knowingly spreading fringe conspiracy theories. Moreover, no platform allows sufficient data access for independent researchers to study the extent of misinformation or rules enforcement. In fact, X and Reddit curtailed data access for researchers in 2023.

Additionally, many platforms rolled back policies against delegitimizing elections in two key areas: rules against claiming early victory and rules against false narratives about past elections. One positive change though was that all four platforms now appear to prohibit threats against election workers, including harassment and threatening violence (a policy that TikTok already had in place).

As of publication, no platform took action on the hateful election misinformation examples ADL reported. ADL also found that the user reporting options in some cases do not match a given platform’s policies. This was especially evident on TikTok and on X.

The report includes detailed recommendations for social media platforms:

• Reinstate prior policies: Rolling back rules against false claims of past fraud or declaring early victory undermines efforts to curb hateful election misinformation. Rumors targeting immigrants and Jews undermine electoral integrity and democracy and put targeted groups at risk, fueling antisemitism and potentially political violence.

• Enforce rules consistently and equitably: Policies are only as good as their enforcement. Platforms must scale the resources necessary to identify and remove prohibited election misinformation, including automated and human review. High profile accounts such as political officials or candidates should not be exempt; on the contrary, because of their influence, they should be held to higher standards.

• Ensure user reporting matches policy: TikTok, X/Twitter, and others must make available to users the complete list of prohibited misinformation through the reporting interface (such as through a link to the full policy details).

• Respond to user reports of election misinformation: ADL’s recent research shows that platforms increasingly do not act on regular user reports, relying instead either on their own proactive detection (often automated) or on reports from reputable organizations like ADL, as part of “trusted flagger” programs. Users can play an important role in preventing the spread of hateful misinformation, especially narratives that require human interpretation to detect, such as those that depend on context or implicit information.

• Apply election rules year-round: Election misinformation does not end after an election; on the contrary, hateful false narratives can grow and develop further, and lead to political violence as they did on January 6, 2021.

• Data access for researchers: Restore (or establish) free or low-cost research APIs for independent researchers that allow for random samples of all content, not just libraries of curated content or top posts.