Inside Iran’s Hidden War: How AI is Reshaping Tehran’s Influence Campaigns

Jerusalem, 19 January, 2026 (TPS-IL) -- As Iran’s regime violently cracks down on unrest at home, Tehran is accelerating a quieter campaign abroad. Experts tell The Press Service of Israel that Iran is increasingly turning to artificial intelligence to supercharge covert influence operations targeting Israel and Western societies, reshaping how disinformation and manipulation are carried out.

For years, Iran has relied on fake online personas, coordinated social media activity, and low-cost psychological pressure. What is changing, analysts warned TPS-IL, is the toolset. Artificial intelligence enables Iranian actors to imitate real people more convincingly, operate across languages fluently, and push narratives at a scale and speed previously difficult to sustain.

The concern is not only technological but strategic. Iran’s expanding use of AI in covert influence campaigns comes amid weeks of internal instability. As the regime suppresses widespread anti-government protests, Tehran is formalizing AI as a national strategic priority. That convergence, analysts warned, could significantly increase the long-term threat to democratic societies.

AI allows Tehran and the regime’s supporters to manufacture images of support or amplify claims when information from inside Iran is tightly restricted. In that environment, AI-driven manipulation becomes harder to verify and easier to weaponize.

“There is no ability today to systematically identify AI-driven influence campaigns,” Danny Citrinowicz, a senior researcher in the Iran and Shi’ite Axis Program at Tel Aviv University’s Institute for National Security Studies, told TPS-IL.

AI as a State Priority

Iran’s interest in AI is not limited to hackers or marginal influence operators. According to analysis published in December by Dr. Avi Davidi of the Jerusalem Institute for Strategy and Security, the drive comes from the top.

As recently as April, Iran’s supreme leader Ayatollah Ali Khamenei described AI as a strategic national issue, warning that future international frameworks could limit access to advanced technologies. He urged Iran to develop what he called the “deep layers” of AI rather than rely on foreign systems.

That directive has been translated into policy. Iran’s parliament approved a National Artificial Intelligence Document outlining state goals, followed by legislation establishing a National Artificial Intelligence Organization under the president’s office. Senior commanders in the Islamic Revolutionary Guard Corps (IRGC) have also spoken openly about using AI in military planning, cyber activity, and decision making.

Davidi’s analysis portrays a regime that sees AI not as a side project but as part of its long-term approach to security and power.

According to Citrinowicz, “The regime uses it to create the impression of broad public support.”

He added that AI manipulates the apparent size of crowds in photos and helps the regime portray protesters as rogue activists seeking to harm Iran. “People are being killed in the streets,” he said. “But when everything is shut down inside Iran, AI makes it easier to convince outside audiences.”

Weaponizing AI

As internal challenges mount, Davidi documents Iranian-linked actors exploiting Western-developed AI tools to enhance cyber and influence operations abroad.

One example is the influence group STORM 2035. According to OpenAI reporting cited in Davidi’s analysis, actors tied to the group used ChatGPT to generate content for fake news sites and social media accounts in multiple languages, including English, Arabic, French, and Spanish. The content focused on polarizing issues such as the Israel–Hamas war, U.S. politics, and social divisions in Western societies, enabling Iran to push messaging that appeared local while concealing its origin.

AI is also used for technical cyber activity. Hackers operating under the CyberAv3ngers persona, linked to the IRGC, used generative AI tools to scan networks, research vulnerabilities, and identify common passwords. OpenAI and Google Cloud reporting cited by Davidi shows the group focused on infrastructure and industrial control systems in Israel and parts of Europe, using AI to speed reconnaissance rather than invent new attack methods.

Another group, APT 42, specializes in social engineering and credential theft. Google Cloud research cited by Davidi indicates the group uses AI to tailor phishing messages, research vulnerabilities, and simulate attack scenarios, improving their ability to impersonate targets across languages and cultures.

OpenAI and Google Cloud did not respond to TPS-IL queries.

From Influence to Pressure

These AI-enabled tactics build on a longer pattern. Over the past several years, Iran has run sustained covert campaigns to shape discourse inside Israel and abroad.

In 2022, researchers exposed an Iranian-linked Facebook network posing as Israelis and Palestinians to spread disinformation. More recently, Israeli security officials warned of suspected Iranian operations targeting the Arab community ahead of elections, as well as bot-driven efforts encouraging graffiti and provocation to inflame internal tensions.

Iranian actors have also been linked to phone hacking and unauthorized access attempts against senior Israeli figures, part of a long-range strategy of pressure rather than isolated cyber incidents.

Experts say AI is accelerating these efforts, allowing Iran to produce more convincing content, operate across languages fluently, and expand its reach to North America and Europe.

Israel’s National Cyber Directorate told TPS-IL that while it operates “around the clock” to address cyber threats, public awareness is critical to prevent the spread of unverified content. “Ultimately, public awareness is the first line of defense against influence operations, with or without AI,” a spokesperson said.

‘War Over Knowledge Itself’

Citrinowicz noted that Iran’s AI-driven influence efforts intensified after Hamas’ October 7, 2023, attack as Tehran sought to weaken Israeli society from within.

AI has reduced gaps that once made Iranian campaigns easier to spot, particularly around language and cultural cues. Deepfakes, avatars, and AI-generated personas are far harder to identify systematically, even as Israel maintains strong cyber defenses.

“Everything we’ve seen, from phone hacking to covert social media campaigns, follows the same logic,” he said. “The goal is to weaken Israeli society internally.”

Israel is due to hold national elections by the end of October, and Citrinowicz warned that Iranian AI operations could severely manipulate the vote. “Artificial intelligence changes the game because it significantly improves Iran’s ability to operate credibly inside Israeli society. This is a very disturbing development,” he stressed.

State Comptroller, Matanyahu Englman has already warned that Israel’s preparations for foreign interference in the elections were “deficient.”

Moran Alaluf, an independent researcher on Iran, described AI as a ״force-multiplier״ for Tehran, enabling it to tailor messages to different audiences at a scale impossible with human operators alone. She warned Iran is likely pursuing independent AI capabilities through cooperation with allies such as China and Russia, while exploiting Western tools in the meantime.

Alex Grinberg, an Iran researcher at the Jerusalem Institute for Strategy and Security, offered a less pessimistic view. While acknowledging Iran’s technological progress, he noted ideological blind spots and a tendency to project a totalitarian mindset onto democratic societies can expose Iranian operations. Mistakes continue to reveal campaigns, he said.

“Iran is moving from basic tactics like phone hacking to advanced technologies, but it still has clear limitations. Their demonization of Israel prevents them from fully understanding Israeli society, often exposing their influence campaigns,” he said.

All experts agreed that the burden increasingly falls on Western societies to verify what they see online.

They also noted Iran’s AI targeting methods are similar in Israel and abroad – aimed at weakening societies internally and tarnishing Israel’s name – though the scale of activity inside Israel is far greater.

In the U.S. and Europe, Iran has used AI to exploit existing political and social divisions. Iranian-linked actors have generated large volumes of tailored content on elections and the Israel–Hamas war, presenting it as local opinion and creating fake profiles to raise funds for pro-Palestinian causes such as the 2025 Gaza flotilla. More recently, Britain’s Daily Telegraph reported a wave of pro-Scottish independence accounts on X abruptly going dark after Iran imposed a internet blackout.

Taken together, the assessments point to a clear trend: Iran’s use of AI in cyber and influence operations may be uneven and opportunistic, but combined with leadership-driven AI development, it creates a growing long-term risk.

Dr. Daniel Cohen, head of the Policy and Technology Program at the Abba Eban Institute at Reichman University, told TPS-IL that Iran’s use of AI goes beyond deepfakes or synthetic avatars.

“This is going to be a war over knowledge itself,” Cohen said. “The Iranians will try to plant messages inside the sources that train AI models. People will then receive disinformation that serves Iranian interests. That is how knowledge gets shaped. They learned this from Russia.”

Preparing for this next phase requires ensuring AI systems are trained on reliable, authentic information. “We need to be ready for this battle over knowledge so that future AI models are grounded in verified sources, not manipulated narratives,” he said.

Cohen added that as countries race to develop AI, Iran’s covert influence campaigns against Israel and the West are likely to become increasingly sophisticated, scalable, and automated.