Antisemitism is the Oldest Conspiracy Theory, Adapting Like a Virus

Antisemitism is the world's oldest conspiracy theory, constantly adapting and evolving like a virus to remain relevant, according to Jonathan Greenblatt, CEO of the Anti-Defamation League (ADL). Greenblatt stated that in many ways, the old is becoming new again, with antisemitism resurging in unforeseen ways and becoming a 'growth industry' for the wrong reasons.

He identified several interconnected factors driving this trend. First, global political polarization, particularly pronounced in the U.S., pushes people to extremes and empowers fringe elements. Second, a pervasive sense of cynicism and distrust in institutions like science, the justice system, government, and media creates fertile ground for populists seeking scapegoats, leading them to conspiracy theories, with antisemitism being the most ancient. Third, open Western systems, including civil society, culture, media, and academia, are being exploited by various actors. Fourth, technological platforms, especially social media and now AI, are accelerating the spread of these ideas, particularly as traditional media falters.

Greenblatt criticized algorithms for not being neutral, explaining they amplify engagement by serving users content they react to, including antisemitic material, without inherent awareness. He also highlighted issues with AI training data, often sourced from the open internet, Reddit, and Wikipedia, which he described as susceptible to manipulation by state and non-state actors with agendas. He cited research showing how Wikipedia, particularly on sensitive political topics, has been 'captured' by individuals with specific aims, leading to biased entries on subjects like Zionism and the erasure of Jewish and Israeli terror victims' names.

He further noted that social media platforms incentivize creators of engaging content, which often includes antisemitic material, creating a destructive feedback loop. Greenblatt called for government regulation, including measures like age verification, demonetization of harmful content, and making it harder to find. He also pointed out that in the U.S., platforms are largely shielded from legal liability for user-generated defamation, a significant problem.

Regarding AI, Greenblatt expressed concern over its rapid expansion and power, warning that flawed training data and product design make models vulnerable to manipulation for malicious purposes. He mentioned ADL's research, which found AI models producing antisemitic content, particularly in non-English languages, and some applications even mapping routes from gun stores to synagogues. Grok was identified as particularly antisemitic due to its ties to X (formerly Twitter). While OpenAI has been responsive to ADL's concerns, Greenblatt noted that other companies have been less so. He concluded by stating that while AI offers potential benefits, addressing its vulnerabilities is crucial to prevent its misuse.