AI language translation was one of the most beneficial uses of AI during major global elections in 2024, and AI misinformation was not as catastrophic as feared
The most significant impact turned out to be AI being used to bridge language gaps everywhere from India to New York and make candidates messages multilingual Eric Smalley / @ericsmalley : AI is now part and parcel of elections, from helping with mundane campaign functions to enabling politicians to speak to constituents in multiple languages at once. Deepfakes, however, were not a big part of the disinformation campaigns in 2024. Bruce Schneier and Nathan Sanders break it down.
Context & Ripple Effects
Pre-election coverage centered on AI’s potential to distort voting, even as India’s parties were already using the technology chiefly for outreach. This assessment shifts the emphasis toward practical multilingual communication, consistent with the earlier constructive campaign use in India.
The result also qualifies the alarm around synthetic political content: Meta reported AI content represented less than 1% of its fact-checked election misinformation, while U.S. research found amplification without evidence of changed voter minds in election disinformation research.
First-order effects
- Campaigns can use translation tools to deliver candidate messages across language communities more quickly, making multilingual outreach a routine campaign capability rather than a specialist service.
- Deepfakes remain a governance concern, but the reported 2024 experience puts immediate attention on ordinary AI-enabled campaign operations rather than a realized catastrophic misinformation scenario.
Second-order effects
- Political consultants and campaign technology providers face pressure to add reliable translation, localization, and review workflows as multilingual reach becomes easier to procure.
- Platforms, fact-checkers, and election officials can calibrate monitoring around observed harms; Meta’s low share of fact-checked AI election misinformation supports separating synthetic-content volume from demonstrated impact.
Third-order effects
- If campaigns continue adopting translation first, AI’s electoral effect may be less about exceptional viral manipulation and more about lowering the operating cost of reaching fragmented language audiences.
- The policy challenge shifts toward dual-use governance: preserving accessibility benefits while maintaining provenance, accountability, and safeguards for deceptive synthetic media.
The trend: Election AI is moving from a feared disinformation shock toward operational infrastructure for campaign distribution, with risk management increasingly tied to measured harms rather than novelty alone.