Since the first publicly available large language model (LLM) ChatGPT hit the scene in November of 2022, the tech has exploded. AI has infested search engines, social media platforms, customer service portals, job interviews, and news articles.
Despite being marketed as “artificial intelligence,” LLMs are decidedly not intelligent on their own, but rather crafty algorithms made to look smart, thanks to clever engineering and no small amount of your data.
Evidently that’s a lesson Yale administrators have yet to learn, as demonstrated by the suspension of international law scholar Helyeh Doutaghi last week based on the whims of an LLM.
Doutaghi, an outspoken advocate of Palestinian rights, began her job as deputy director of Yale’s Law and Political Economy project in 2023. Her work was cut short when the university’s admin banned her from campus and put her on administrative leave, following accusations that she was a “terrorist” by Jewish Onliner, a far-right blog that describes itself as “empowered by AI capabilities.”
In a recent investigation into far-right AI bots, the Israeli newspaper Haaretz recently noted that Jewish Onliner has been linked to Israeli disinformation campaigns, using LLM-powered bot accounts to troll and harass users whose posts it disagreed with.
In this case, the blog had accused Doutaghi of membership in Samidoun, an international advocacy group organizing to raise awareness of Palestinian political prisoners. Though Israel has designated Samidoun a terrorist group — an accusation it shares with organizations like Defense for Children International and the Addameer Prisoner Support and Human Rights Association — the US has not.
Regardless, the AI-powered blog’s accusation noted the scholar’s speaking engagement on panels it alleged were sponsored by Samidoun, indicting her as a member of the group. Doutaghi, for her part, denied that she was a member in a statement on Mondoweiss.
“I was given only a few hours’ notice by the administration to attend an interrogation based on far-right AI-generated allegations against me,” the statement reads, “while enduring a flood of online harassment, death threats, and abuse by Zionist trolls… Just a few hours later, [Yale] placed me on leave, revoked my IT access — including email — and banned me from campus.”
The swiftness with which AI-powered allegations were leveraged against Doutaghi is raising concern for free speech, especially in light of Donald Trump’s executive order targeting antisemitism, which is seen by some as a gateway to criminalizing pro-Palestinian speech.
It also comes as pro-Palestinian activist Mahmoud Khalil was hauled out of his Columbia University dorm by plainclothes ICE agents last Saturday, as part of Secretary of State Marco Rubio’s “Catch and Revoke” campaign, which uses AI to scrub social media for the expression of anti-Israel sentiment.
Despite not being charged with a crime, Khalil — a lawful permanent resident of the US — was ripped from his pregnant wife and whisked to a federal detention facility in Louisiana, leaving his whereabouts unknown to his family and attorneys for over 24 hours.
Failing to declare formal charges or evidence of criminal activity, a spokesperson for the Department of Homeland Security said that Khalil was arrested “in support of President Trump’s executive orders prohibiting antisemitism.”
It’s a bone-chilling precedent for free speech in the US, harkening back to the second red scare, when union organizers, peace activists, and civil rights leaders were forced out of their jobs and stalked by law enforcement on the basis of ideology alone.
The difference? This time, AI’s helping call the shots.
More on AI: Police Department Testing AI-Powered Detective on Real Crimes