Israel's Elections Committee Meets Tech Giants Over AI Threats

Israel's Central Elections Committee, led by Chair Noam Solberg, convened with representatives from major technology companies including Meta, Google, Microsoft, and TikTok at the President's Residence. The meeting, also attended by President Yitzhak Herzog, focused on concerns regarding the potential misuse of artificial intelligence (AI) for election manipulation, including deepfakes, disinformation, and influence operations.

President Herzog emphasized the "real dangers" posed by AI in elections, particularly in the context of the upcoming vote being the first since October 7th. He highlighted the critical role of tech platforms in safeguarding election integrity and urged the public to be vigilant against false information. Solberg described social media as the "new town square" where citizens receive information and engage in political discourse, stressing the need to protect election fairness while upholding free speech.

The meeting aimed to deepen cooperation between the Elections Committee and the tech platforms, acknowledging the committee's limited authority in the online sphere. Companies presented their respective policies, technological capabilities, and ongoing efforts to identify and address AI-generated content and coordinated influence campaigns. TikTok, for instance, prohibits paid political advertising and requires labeling of AI-generated content, while Google allows election ads subject to advertiser verification and political advertising rules.

This gathering follows months of preparation, including the establishment of a joint task force by the Elections Committee and the Shin Bet earlier in the year to counter foreign influence attempts. A recent exercise involving the Elections Committee, police, and Shin Bet simulated cyber threats, foreign influence operations, and AI misuse, revealing challenges in identifying coordinated activity in real-time. Israel's current election cycle is the first to implement a requirement for labeling propaganda that uses digital means to create or alter realistic-looking images, events, or documents, a rule already invoked when Solberg ordered the removal of an AI-generated campaign video depicting a fabricated event.