Israeli Researchers Launch AI Monitoring Project for Election Campaign Content
Israel's 26th Knesset election marks the first national vote conducted amid the rise of artificial intelligence (AI) technologies. In response to new challenges and concerns, a research team led by Professor Anat Ben-David from the Open University’s Department of Sociology, Political Science, and Communication has established the initiative "This Is AI." The project monitors social media accounts of political parties and candidates to map AI usage in election campaigns, ranging from adding subtitles to creating simulated events that never occurred. Suspicious content undergoes specialized AI detection tools and rigorous human verification before public disclosure.
Starting this Sunday, the initiative will be featured in a new segment on Globes’ "The AI Behind the Elections," providing weekly updates on AI trends in campaign propaganda. The verification process includes automated collection of posts and videos, initial technological analysis for AI indicators, double human review by independent checkers, and transparent publication on a public dashboard. This system aims to clarify how AI is used without serving as a regulatory authority.
An example of AI use in campaigns is a video released by Prime Minister Benjamin Netanyahu depicting Avigdor Lieberman and Gadi Eizenkot riding a Trojan horse toward government gates, with Yair Golan and Mansour Abbas hidden inside. While clearly a caricature, it illustrates how AI tools have expanded campaign methods and blurred lines between dramatization and realistic footage. Professor Ben-David notes that AI can also enhance videos with jingles, subtitles, or visual improvements, but it can also fabricate scenes or statements that never happened.
In July, Israel amended election laws to require clear labeling of "deepfake" content, visual or audio material digitally created or altered to appear authentic. However, not all AI-generated content requires disclosure, such as added subtitles or jingles. The "This Is AI" project tracks all AI uses, identifying patterns and users, and publishes findings with detailed data by party and candidate. Currently, about 66% of monitored campaign content consists of videos, with AI usage still relatively low but growing.
The researchers emphasize that the main concern is not just isolated fake videos but how AI-driven content proliferation changes political discourse and voter trust. Funded by the Meir Shamgar Center for Digital Law and Innovation and the Edmond J. and Lily Safra Center for Ethics at Tel Aviv University, with support from Bright Data’s Bright Initiative, the project aims to provide transparency and help the public and journalists navigate the evolving digital campaign landscape.
