In preparation for a special session held by the Israeli Parliament’s Science and Technology Committee on December 25, 2023, and at the behest of the Parliamentary Research and Information Center, the Israel Internet Association (ISOC-IL) presented the Committee with a comprehensive brief and supporting materials on the employment of AI in wartime disinformation campaigns.
This comprehensive brief proved to be a cornerstone in this special session and was extensively cited in the documents prepared by the Parliamentary Research and Information Center leading up to it. Prior ISOC-IL publications recommending a framework for the regulation of social networks in Israel were cited as well.
Summary of the ISOC-IL Brief:
For almost a decade, ISOC-IL’s Internet Safety Hotline has provided the Israeli public with personal assistance in dealing with online abuse, threats and offensive content – especially on social media. In the wake of the October 7 massacre carried out by Hamas on Israeli civilians and Israel’s military response in turn, our hotline immediately responded to an unprecedented surge in public complaints regarding false information and offensive social media content related to the war. We received and reported numerous instances of false content, justification of terrorism, and hate speech against Israel and Jews across various messaging apps and platforms. We also published vital emergency guides to help people avoid the graphic content and cyberattacks that have been a significant element of the war being waged on Israel.
One of the most concerning aspects of the current rush of false content disseminated online is the deliberate, widespread weaponization of artificial intelligence to create posts, profiles, images and other materials with the sole aim of promoting anti-Israel narratives. AI-synthesized content comes in the form of text, audio, images or video and can even include profiles impersonating humans. Advanced generative AI enables foreign actors to create fake localized Hebrew content that is more sophisticated and difficult to detect, including manipulated graphic content. Our researchers have mapped the different forms of fabricated content into seven known categories of mis/disinformation commonly proliferated online during conflicts and emergencies.
Our research demonstrates that foreign networks, in an effort to permeate Israeli discourse and sow division and panic among the population, have been using partially AI-powered coordinated inauthentic behavior (CIB) to target Israelis with waves of inflammatory content. This method exploits AI automation not to improve the quality of the deception, but rather to amplify the reach and volume of the false content – evading detection by overwhelming the various platforms’ moderation mechanisms.
Though it was intended to fracture societal resilience during fighting, this coordinated influence has exposed a systemic unpreparedness and legislative gaps in identifying and combating AI-powered online threats, and ISOC-IL is proud to play a part in building Israel’s protective system.
***
The Full Report (in Hebrew) is available here. For additional information and collaborations, contact our International Department at [email protected].
The post Weaponizing AI for Disinformation: An ISOC-IL Brief to Israeli Lawmakers appeared first on Israel Internet Association – (ISOC-IL).