Heads of state, top government officials, leaders of international organizations, tech CEOs, and academics from over 90 countries are meeting this week in Paris for the Artificial Intelligence Action Summit. In his speech at the summit, Vice President JD Vance, who led the U.S. delegation, stated: "We will safeguard American AI and chip technologies from theft and misuse, work with our allies and partners to strengthen and extend these protections and close pathways to adversaries attaining AI capabilities that threaten all of our people."
One such adversary not invited, for good reason, was Iran. In its approach to AI, Iran is following its successful development of drone technology, which began with the reverse engineering of a U.S. RQ-170 Sentinel drone it downed in 2012 by using its technology to produce its Shaheed drones. As it now develops its AI expertise via theft, copying open-source software, and collaborating with Russia and China, the Iranian government – which recently acknowledged its failure to reverse-engineer a Tesla electric vehicle – is quietly pursuing AI advances for military, cyber, and defense purposes.
The world has already witnessed the dangers of Iran's drone program, and Iran's development and destructive capabilities have surprised Western military and intelligence agencies. If proper attention is not given to another nascent Iranian threat now emerging: a clandestine yet intensifying rush for mastery of AI – this too will become another national security threat.
Iran's investments into its AI engineers seem to be paying off. So far in 2025, Iran has announced the following: it is allocating $115 million to AI research, it has added an "advanced data-processing" warship equipped with AI to its naval fleet and 1,000 new combat drones, some with AI navigation, to its army; and it has boosted the precision of its missiles using AI, and many of these new weapons will be stationed as part of electronic warfare units in the strategic port of Jask on the Gulf of Oman. It also announced it is testing 'smart' combat robots.
According to findings released by Google on January 29, 2025, Iranian groups are using the company's Gemini AI chatbot for researching defense organizations to target with hacking attempts, to generate content in English, Hebrew, and Farsi to be used in phishing campaigns, and other purposes.
For Iran, AI is an inexpensive path to power. It is already using it for cyberattacks and influence operations abroad, and to upgrade its drones and ballistic missiles, deploy swarms of unmanned watercraft, surveil and repress its citizens, and control its borders. Alarmingly, it could now be exploring AI's potential nuclear applications.
The U.S. government's October 2024 landmark national security memorandum on AI reflects awareness of such threats, and details how the U.S. "must develop safeguards for its use of AI tools, and take an active role in steering global AI norms and institutions." The memorandum sets out guidelines for the Pentagon, intelligence agencies, and national security organizations to add "guardrails" to prevent AI from being hijacked by hostile governments such as Iran, to develop nuclear weapons or "generate or exacerbate deliberate chemical and biological threats." It also calls on intelligence agencies to begin protecting work, and the chips used to power it, as national treasures.
The Iranian government has already conducted successful campaigns using AI. The Iranian covert influence operation known as Storm-2035 used ChatGPT to disseminate disinformation and generate false news reports on a variety of topics, including the U.S. presidential election.
On numerous occasions, the Iranian leadership has declared its intent to make the country an AI superpower. By November 2021, it completed an official "study of the national artificial intelligence development roadmap," and Iranian Supreme Leader Ali Khamenei has since then publicly discussed Iran's strategy to master AI. He said that "one of Iran's goals is to be among the top 10 countries in the area of artificial intelligence," and asked Iranian scientists to focus on this goal.
Less than two months after Iran's launch of its national AI center, Khamenei stressed, in August this year, that Iran must act fast to "master all layers of AI." He went on to warn of "attempts by the world's opportunists and those seeking power to target Iran by "establish[ing] an IAEA-like body [to regulate] AI" and "prevent countries' progress in this field."
Based on their ongoing public cooperation, it should be deduced that Russia and China have already provided AI technology and knowledge to Iran. In 2020, the Iran-Russia Cooperation Group on Communications and Information Technology began working together on AI, and this year, the two countries signed an official agreement on AI cooperation. Iranian Army deputy chief of staff Brig.-Gen. Mehdi Rabbani attended a Beijing forum on topics including AI back in 2019. By January 2022, the vice chancellor of a top Iranian university was pointing at the 3,000 Iranian students already in China and stressing that the 25-year Iran-China agreement of March 2021 offered a "golden opportunity" to boost the two countries' AI cooperation.
The most dangerous potential of Iranian mastery of AI is for military uses and nuclear weapons technology. Ali Akbar Salehi, head of the Atomic Energy Organization of Iran (AEOI), said in January 2021 that the organization was striving to use the quantum technology it had been developing since 2016 for AI.
The IRGC and Iran Army added to their arsenals in July 2023 an Iranian-made missile that, Iran claimed, "has artificial intelligence and stealth capabilities." IRGC commander Gen. Hossein Salami said in February 2024 that Iranian missiles and vessels were "equipped with highly accurate AI technologies," adding that "today, we can move from [the Iranian nuclear facility at] Bushehr to the American coasts." IRGC Aerospace Force commander Brig.-Gen. Amir Ali Hajizadeh boasted, in January 2021, about "the use of artificial intelligence technology" in Iran's surface-to-surface ballistic missiles. Just yesterday, on February 10, IRGC Naval Commander Ali Reza Tangsiri announced the unveiling of supersonic cruise missiles with a range of 2,000 kilometers, saying: "All these missiles are sophisticated – designed and manufactured on the basis of artificial intelligence – and will be unveiled next year."
The October 2024 U.S. security memorandum gives the U.S.'s new AI Safety Institute (AISI) the power to help inspect AI tools prior to release to make sure that they cannot be used by terrorists and bad actors to build weapons of mass destruction or help hostile nations improve missile accuracy, as Iran admits doing. Amid reports in 2023 that the IRGC navy had developed AI-enhanced unmanned military vessels, that could be used for mass attacks, commander Tangsiri said that the navy had developed AI-equipped long-range strategic cruise missiles. In August, IRGC naval exercises featured AI-enhanced unmanned vessels. Tangsiri explained that the IRGC had fitted drones with AI for improved range, precision, and radar evasion, and that the IRGC was implementing AI in the drone, missile, vessel and submarine industries.
As AI development is moving at the speed of light in Iran, Khamenei is right to worry about international efforts to stop Iran from using it for advancing its military aims. But these efforts, like efforts to slow Iran's nuclear program, will be difficult to bring to fruition. They will need to include the usual measures, such as leveling sanctions any time the country is caught trying to obtain technology by illegal means.
The new Trump administration will need to act fast to strengthen guardrails to ensure that America's enemies such as Iran are not able to access its technology. The severity of the threat was reflected in statements by the head of the Pentagon's newly created Artificial Intelligence Office, who acknowledged in December that "AI adoption by adversaries like China, Russia, Iran, and North Korea is accelerating and poses significant national security risks." This week China surprised the world with its DeepSeek AI, action should be taken to make sure Iran doesn't have the same opportunity with its own AI model.
*Steven Stalinsky, Ph.D., is Executive Director of The Middle East Media Research Institute.
