Skip to content
Submit incident
AII-1002-46231
ReportedReported by fewer than three independent publishers, with no official finding.

Meta's AI moderation fails to flag election misinformation on Facebook

Meta's moderation systems failed to flag recurring election misinformation on Facebook ahead of the 2024 U.S. presidential election, even though the same false voting instructions had been flagged during the 2020 election. Election officials in multiple states reported that the misinformation spread widely without any Meta intervention until state and local authorities issued their own warnings.

July 2024 Β· Meta Β· Misinformation and synthetic media
July 2024
AII-1002-46231View source β†—
Quick facts
Category
Misinformation and synthetic media
Country
United States
Source
CNBC
Date
July 2024
Drafted by
Raidu Incident Detection

What happened

A Facebook post containing false voting instructions circulated widely in July 2024, advising voters to request new ballots if poll workers wrote anything on their ballot forms because it would be invalidated. The same misinformation had spread on Facebook during the 2020 election, when Meta flagged it as false information with a link to a USA Today fact-check.

In 2024, the same post spread without any Meta moderation label or fact-check link. The misinformation was widespread enough that the North Carolina State Board of Elections issued a press release on August 2 warning that such posts had been circulating for years and had resurfaced in multiple counties. Multiple Facebook users in North Carolina, Mississippi and New Jersey continued sharing the ballot misinformation without any notification that it was false.

Reported impact

Derek Bowens, director of elections in Durham County, North Carolina, said he faced a potential crisis when the misinformation began spreading virally. Election officials across the U.S. expressed concern about Meta's ability to control misinformation as the 2024 presidential election approached, citing the spread of false claims that threatened to disrupt voting.

Index analysis

Meta had reduced its trust and safety and customer service staffing through layoffs and deprioritized news content, making it difficult for election officials to dispel misinformation. When CNBC flagged posts with the false information to Meta, a company spokesperson said the posts had been sent to third-party fact-checkers for review, but no moderation action was taken before state and local officials intervened.

Classification

Category
Misinformation and synthetic media
Industry
Media, social platforms and marketing
Country
United States
Harm status
Potential harm
Intent
Unknown
Timing
Post-deployment
Harm types
Public interest
Harmed
General public, Government
Failure modes
Content moderation failure
AI systems
Facebook AI Moderation System

Sources and evidence

This record was researched and written by the Index, based on the following original reporting.

CNBC
Meta's misinformation problem has local election officials struggling to get out the truth
26 September 2024