Tehran is reportedly utilizing artificial intelligence to conduct surveillance on Jewish communities and Iranian dissidents worldwide, according to evidence presented during an Australian parliamentary inquiry examining AI risks. The move has raised concerns about foreign interference and digital coercion targeting diaspora populations, particularly within Australia.
Rana Dadpour, director of Australian United Solidarity for Iran, submitted warnings to the parliamentary joint select committee on artificial intelligence about the growing use of AI by authoritarian regimes to extend their reach beyond national borders. She emphasized that foreign states, including Iran, are harnessing AI technologies to monitor, intimidate, and suppress political activism among diaspora groups. Dr. Dadpour urged improved information sharing between leading AI developers—such as Google, Anthropic, OpenAI, and Microsoft—and Australia’s newly established AI Safety Institute, which currently operates with a staff of 35.
“Surveillance, hacking, online harassment and related practices can induce self-censorship, disrupt diaspora networks and constrain political participation in democratic host countries,” Dr. Dadpour noted in her submission. She warned that such digitally enabled coercion can prevent individuals from exercising fundamental freedoms, including free expression and association.
In a separate disclosure, AI company Anthropic revealed that in September it uncovered Iran-linked actors using its Claude chatbot to automate profiling and identity harvesting of Jewish individuals in Israel and other countries. Anthropic reported shutting down 16 Claude accounts linked to Iranian paramilitary and domestic intelligence agencies deploying AI for surveillance and social media analysis.
The inquiry also heard from Australia’s Special Envoy to Combat Islamophobia, Aftab Malik, who raised concerns about AI systems manifesting biases against Muslims. Malik called on the government to mandate watermarking or clear labelling of AI-generated content and to enforce pre-implementation assessments examining bias, discrimination, and human rights implications. He highlighted how algorithms designed to maximize engagement often amplify hate speech targeting Muslim communities.
Gender-based impacts of AI were a key focus during the inquiry as well. Submissions from advocacy organizations underscored risks of AI reproducing and exacerbating violence against women and children through deepfakes, doxxing, impersonation, stalking, and cyber harassment. Our Watch, a gender equality organisation, warned that AI-powered geolocation technologies could be used to track women and children fleeing domestic violence. They also noted harmful gender norms perpetuated online, influencing young people’s attitudes toward abuse.
Child safety organisations joined calls for stronger regulatory frameworks. The Alannah and Madeline Foundation urged the government to incorporate AI into proposed “duty of care” legislation for technology companies and to embed “safety by design” principles in AI development. CEO Sarah Davies highlighted the importance of addressing risks not just through consumer caution but through design and operational safeguards, drawing lessons from past online harms suffered by children.
Australia’s eSafety Commissioner, Julie Inman Grant, reported that three-quarters of children surveyed have used AI assistants, underscoring the urgency of addressing AI’s societal and safety impacts. The parliamentary committee continues to evaluate these issues as it develops recommendations on the responsible deployment and regulation of AI technologies.