Meta: Only 'Modest' AI Impact on World Elections
Social media company Meta has reported that artificial intelligence-generated content related to elections, politics, and social topics accounted for less than 1% of fact-checked misinformation on its platforms including Facebook and Instagram....
0:00
/1861
Facts
- Social media company Meta has reported that artificial intelligence-generated content related to elections, politics, and social topics accounted for less than 1% of fact-checked misinformation on its platforms including Facebook and Instagram.[1][2]
- In a blog post on Tuesday, Global Affairs chief Nick Clegg claimed that, in 2024, Meta tracked elections in the US, EU, UK, India, Mexico, and Brazil among others — deploying content policy updates, an updated penalties system, and voting alert notifications.[3][4]
- Meta's top-of-feed reminders on platforms like Facebook and Instagram received over 1B impressions during the US election, providing information on voter registration, methods of voting, and reminders to vote.[5]
- Clegg said Meta's AI image generator, Imagine, blocked 590K requests to create images of US Pres. Joe Biden and both Republican and Democratic presidential tickets in the month leading up to the US election. Clegg also claims Meta took down 20 covert operations globally during the election period.[6][7]
- Meta also claimed that identified groups engaged in coordinated inauthentic behavior (CIB) did not solely work on Meta's platforms but also other social media sites including YouTube, TikTok, X, Telegram, Reddit, Medium, Pinterest, and more.[8]
- Meta alleged that Russia, Iran, and China were the primary sources of covert influence campaigns. It also acknowledged that 'harmless' content was restricted or taken down 'too often,' with Clegg telling reporters that, in hindsight, Meta 'overdid' content moderation during COVID.[9][3]
Sources: [1]GovTech, [2]Reuters, [3]Meta, [4]Euronews, [5]PetaPixel, [6]Engadget, [7]Washingtontimes, [8]DIGIT and [9]Verge.
Narratives
- Narrative A, as provided by MIT Technology Review. The world’s democratic elections in 2024 were overshadowed by dire warnings of AI-driven interference, from deepfakes to microtargeted manipulation. Evidently, these fears were exaggerated. While AI’s misuse warrants vigilance, the obsession with its alleged dangers risks overshadowing other, far deeper threats to democracy.
- Narrative B, as provided by Brennan Center for Justice. While AI's impact on recent elections may have been limited, its growing use — ranging from deceptive deepfakes to creative campaign tools — underscores a need for vigilance. Policymakers must safeguard systems without stifling legitimate expression. Transparency, rapid responses, and robust oversight are key, and complacency is not an option.
