The Dead Scientists Problem: Iran's Shattered Nuclear Program and the Black Market Threat No One Can Track

At 3:21 a.m. on June 13, 2025, Israel launched Operation Narnia — a simultaneous assassination campaign that killed nine of Iran's top ten nuclear scientists in their beds, using a classified weapon whose nature remains under military censorship [1]. In the days that followed, additional strikes brought the total to at least 15 scientists killed during the 12-Day War [2], part of a broader pattern of targeted killings stretching back to 2010 that has now claimed more than 20 Iranian nuclear researchers [3].

The tactical achievement was immediate. But the strategic aftermath is far murkier. Iran's enriched uranium stockpile — nearly 10,000 kg as of mid-June 2025, including 441 kg enriched to 60% U-235 — has been unverifiable by the International Atomic Energy Agency since the strikes [4]. Thousands of lower-ranking scientists, engineers, and technicians remain alive and scattered. And the international safeguards system designed to prevent nuclear materials and expertise from reaching the black market has, by the IAEA's own admission, significant gaps it cannot close [5].

The Kill List: 15 Years of Targeting Iran's Nuclear Brain

The campaign to eliminate Iranian nuclear scientists began in January 2010, when Masoud Ali-Mohammadi, a University of Tehran physics professor, was killed by a remote-controlled motorcycle bomb [3]. Over the next two years, three more scientists were assassinated — Majid Shahriari and Darioush Rezaeinejad by motorcycle-riding gunmen, and Mostafa Ahmadi Roshan by a magnetic car bomb — while a fifth, Fereydoon Abbasi, survived a similar attack [3].

The most consequential pre-2025 killing came on November 27, 2020, when Mohsen Fakhrizadeh — widely considered the architect of Iran's nuclear weapons research — was shot dead by a remote-controlled machine gun outside Tehran [6]. Iran blamed Israel. No government has officially claimed responsibility for any of the 2010–2020 killings, though intelligence officials and media investigations have attributed them to Israel's Mossad, sometimes with logistical support from the CIA [7].

Source: Multiple sources; Wikipedia, FDD, IAEA

Data as of Jun 30, 2025CSV

Operation Narnia in June 2025 represented a qualitative escalation. Rather than covert, deniable operations, Israel openly targeted scientists as part of a declared military campaign. Those killed included Mohammad Mehdi Tehranchi, one of Iran's most prominent physicists; Fereydoon Abbasi, the survivor of the 2010 attack who had since become head of Iran's Atomic Energy Organization; Hossein Jabal Amelian, chairman of Iran's Organization for Defensive Innovation and Research (SPND); and Akbar Motallebizadeh, Said Borji, and several others linked to weaponization research [1][2][8]. At least 71 civilians also died in the strikes, including children [1].

What's Unaccounted For: Uranium and Centrifuges

As of June 13, 2025, the IAEA estimated Iran's total enriched uranium stockpile at 9,874.9 kg [4]. The breakdown by enrichment level tells a specific story about weapons proximity:

Source: IAEA Reports

Data as of Jun 13, 2025CSV

The 441 kg enriched to 60% is the most strategically significant quantity. According to the Center for Arms Control and Non-Proliferation, this material represents roughly half of Iran's total enrichment effort measured in separative work units (SWU), and converting it to 90% weapons-grade would require only 564 additional SWU — about 1% of the work already completed [9]. A single cascade of IR-6 centrifuges could produce enough weapons-grade material for one device every 25 days [9].

Since the June 2025 strikes, the IAEA has been unable to verify Iran's full stockpile. The agency's September 2025 report noted "near zero IAEA oversight of Iran's nuclear programme," with the entire enriched uranium stockpile unaccounted for since mid-June [5]. Iran allowed inspectors to return to undamaged facilities like the Bushehr reactor but denied access to damaged sites and provided no information about where the enriched uranium stockpile was moved [5].

This verification gap did not begin in 2025. Iran stopped implementing the Additional Protocol to its safeguards agreement in February 2021, following the Fakhrizadeh assassination and the passage of a parliamentary law restricting inspections [10]. The IAEA has acknowledged that it lost "continuity of knowledge" regarding centrifuge production, heavy water inventory, and uranium ore concentrate — a loss it described as irreversible [4].

Separately, approximately 10 kg of nuclear material previously assessed by the IAEA as available to Iran's Applied Physics Institute at Lavisan-Shian remains unaccounted for, with the agency stating its current whereabouts are unknown [4].

The Price of Nuclear Expertise on the Black Market

The A.Q. Khan network — the most extensively documented nuclear black market in history — provides the closest benchmark for pricing. Khan's operation, which ran from the 1980s through 2003, offered a menu of services: centrifuge designs, enrichment plant blueprints, and even a Chinese-origin nuclear warhead design, sold to Iran, Libya, and North Korea through intermediaries in over 20 countries [11].

Specific transaction values are partially documented. A 1998 letter released by Khan himself referenced a $3 million transfer from North Korea to Pakistani military officials in exchange for centrifuges and technical expertise [12]. Libya's more comprehensive package — including centrifuge components, UF6 feedstock, and a weapons design — was valued by investigators at approximately $100 million [11]. Khan's network functioned as what the International Institute for Strategic Studies called "a virtual one-stop-shop" for nuclear weapons capability [13].

The expertise component — the human knowledge of implosion lens design, neutron initiator construction, and enrichment cascade operation — is harder to price in isolation. But the Khan precedent demonstrates that a single knowledgeable individual, willing to sell, can catalyze an entire national weapons program. North Korea obtained centrifuge designs and operational knowledge from Pakistan between roughly 1990 and 2000; by 2006, it had conducted its first nuclear test [14].

The Brain Drain Precedent: Lessons from the Soviet Collapse

The closest historical parallel to the current situation is the post-Soviet nuclear brain drain of the 1990s. When the USSR dissolved, approximately 60,000 specialists involved in weapons of mass destruction development — including an estimated 20,000 in the nuclear field — faced economic collapse [15]. Salaries went unpaid for months. In 1996, the director of Chelyabinsk-70, one of Russia's two nuclear weapon design centers, committed suicide over his inability to pay staff [15].

Western governments treated this as an acute proliferation emergency. The International Science and Technology Center (ISTC), established in Moscow in 1992, spent nearly $190 million on research grants designed to keep former weapons scientists employed in civilian work [15]. The Nunn-Lugar Cooperative Threat Reduction program supplemented these efforts. The fear — that desperate scientists would sell their knowledge to states like Iran, North Korea, or Libya — was never fully realized at scale, but individual cases of attempted recruitment by foreign intelligence services were documented [15].

The Iranian situation differs in key respects. Iran's nuclear workforce is smaller than the Soviet arsenal complex. But the instability is arguably greater: the scientists being targeted are not facing mere economic hardship but active military operations against their workplace. And the international community has not established any equivalent of the ISTC to offer alternative employment or monitor the movements of displaced Iranian nuclear personnel.

Who Would Recruit Them?

Several state and non-state actors possess both the financial resources and strategic motivation to acquire Iranian nuclear expertise.

North Korea, which built its uranium enrichment program on Khan network technology, has demonstrated willingness to purchase nuclear know-how from foreign scientists [14]. Saudi Arabia has publicly stated its intention to match any Iranian nuclear capability and has the financial means to recruit [16]. Non-state actors face higher barriers — Kelsey Davenport of the Arms Control Association has noted that non-state groups "lack infrastructure for weapons-grade enrichment" — but she warned that "if the current Iranian government implodes or the conflict causes significant internal instability, there is an increased risk that nuclear materials will be stolen or diverted" [8].

The CIA's own approach underscores the perceived value of these scientists. A New Yorker investigation traced a CIA defector recruitment campaign run by case officer Kevin Chalker, whose pitch to Iranian scientists was blunt: "We can debrief them and learn so much more — and, if they say no, then you can kill them" [17]. Three-quarters of approached scientists reportedly agreed to cooperate [17]. Intelligence gathered from these defectors contributed to the Stuxnet cyberattack, the 2015 JCPOA negotiations, and targeting data for the 2025 strikes [17].

The Counterproductive Case: Are Targeted Killings Making Things Worse?

The strongest argument against the assassination campaign rests on four claims, each supported by historical evidence.

First, killing known scientists eliminates intelligence assets. Fakhrizadeh was someone the IAEA had sought to interview for years. His death removed any possibility of his eventual cooperation — voluntary or coerced — and destroyed a node in the intelligence picture of Iran's program [18]. The Bulletin of the Atomic Scientists argued that the assassination "eliminates a valuable source of intelligence" and that his surveillance value alive exceeded his removal value dead [18].

Second, assassinations push programs underground. After the 2010–2012 killings, Iran decentralized its nuclear infrastructure and increased operational security. After Fakhrizadeh's death, Iran's parliament passed legislation restricting IAEA inspections, directly reducing international visibility into the program [10]. Countries can now argue that revealing the identities and roles of nuclear personnel makes those individuals targets — a rational response that fundamentally undermines the IAEA's verification methodology, which depends on access to key personnel [10].

Third, killing scientists scatters expertise beyond traceable networks. Jim Lamson, a former CIA analyst now at the Center for Nonproliferation Studies, acknowledged that "there will be replacements of the managers and scientists, but the impact on the killed officials' experience and expertise will be hard to replace." He added a critical caveat: "It is always possible that Iran has additional sites that were not known to Israel and the U.S." [8].

Fourth, the campaign may strengthen rather than weaken resolve. Research from Georgia Tech's Sam Nunn School of International Affairs found that targeting campaigns may "reinforce domestic support for a government, which could then redouble efforts toward nuclear development" [19]. The researchers concluded that "targeting alone is unlikely to destroy a program outright" and characterized it as "a risky option that may fail to disrupt an enemy nuclear program while sparking public outrage" [19].

Defenders of the campaign counter that Iran was months from weapons capability and that diplomatic alternatives had been exhausted. Israel's military described Operation Narnia as "a significant blow to the regime's ability to acquire weapons of mass destruction" [19]. The practical question — whether removing 15–20 senior scientists genuinely degrades a national program when thousands of trained engineers remain — does not have a clear historical answer.

The Legal and Monitoring Vacuum

UN Security Council Resolution 1540, adopted in 2004, obligates all states to adopt domestic legislation preventing the proliferation of nuclear, chemical, and biological weapons to non-state actors [20]. But the resolution focuses on materials and delivery systems, not on monitoring the employment of individual scientists. No international legal framework specifically tracks or restricts where former nuclear weapons researchers work after leaving a program.

The Budapest Convention on Cybercrime, sometimes cited in this context, addresses digital offenses and has no direct application to nuclear expertise transfer [20]. The practical result is that a displaced Iranian enrichment engineer could, in principle, accept employment in any country without triggering any international notification mechanism.

The IAEA's own monitoring capacity has degraded in parallel. Between February 2021, when Iran curtailed inspections, and the June 2025 strikes, the agency steadily lost verification capabilities. As of early 2026, Iran had provided access to only four of six remaining unaffected nuclear facilities [4]. The IAEA has passed resolutions demanding greater access, but enforcement mechanisms remain limited to referral to the UN Security Council, where political dynamics have historically blocked action [5].

Source: OpenAlex

Data as of Jan 1, 2026CSV

Academic research on nuclear proliferation and safeguards surged after 2022, peaking at over 5,300 papers in 2024 — a reflection of growing scholarly concern about the adequacy of existing monitoring frameworks [21].

From Expertise Loss to Capability: How Long Does Reconstitution Take?

Historical cases suggest the timeline from acquiring a recruited expert to achieving meaningful nuclear capability is measured in years, not decades — but depends heavily on pre-existing infrastructure.

Pakistan's program, built partly on centrifuge designs Khan obtained while working in the Netherlands, took roughly 15 years from initial technology acquisition (mid-1970s) to a deliverable weapon (1998) [11]. North Korea's uranium enrichment path, seeded by Khan network transfers beginning around 1990, produced a first nuclear test in 2006 — approximately 16 years [14]. But both programs had dedicated state resources, existing scientific establishments, and decades of sustained effort.

A well-funded non-state actor starting from acquired expertise alone would face substantially longer timelines, given the industrial infrastructure required for enrichment. Davenport noted that non-state actors lack the "infrastructure for weapons-grade enrichment" and would face "significant challenges" even with recruited scientists [8].

The more immediate risk may not be a complete weapons program but rather a "dirty bomb" scenario using diverted radioactive materials, or the sale of specific technical knowledge — such as centrifuge cascade configurations or implosion lens geometry — that could accelerate an existing state program by years.

What Comes Next

As of April 2026, Iran has threatened to withdraw from the Nuclear Non-Proliferation Treaty [22]. The IAEA remains unable to verify the location or integrity of Iran's enriched uranium stockpile. The United States and Israel continue to target Iranian nuclear personnel and infrastructure [23]. And no international mechanism exists to track the movements or employment of the surviving scientists, engineers, and technicians who spent years or decades working on one of the world's most advanced enrichment programs.

The A.Q. Khan network demonstrated that a single motivated individual can seed nuclear programs across continents. The post-Soviet experience showed that economic desperation can overcome even deep institutional loyalty. The question facing the nonproliferation regime today is whether the destruction of Iran's nuclear leadership has contained a threat — or created dozens of smaller, less visible ones.