The Nuclear Question: Did Operation Epic Fury End Iran's Bomb — or Guarantee It?

Did Operation Epic Fury end Iran's bomb — or guarantee it? The historical record answers.

Eleven weeks of bombing degraded Iran's nuclear infrastructure. Whether it ended the program, delayed it, or made an Iranian weapon inevitable is the war's central unanswered question — and the historical record offers an uncomfortable answer.

When the first wave of Operation Epic Fury struck Iran's nuclear facilities on 28 February 2026, the stated objective was unambiguous: to end Iran's path to a nuclear weapon by force. The U.S. component of the joint campaign with Israel dropped fourteen GBU-57 bunker-penetrating bombs on the underground enrichment sites at Fordow and Natanz and destroyed the Isfahan conversion complex with Tomahawk cruise missiles; the opening salvo also killed Supreme Leader Ali Khamenei, whose death Iranian state media confirmed on 28 February [1][2][3]. The strikes were the culmination of a confrontation that had been building since the June 2025 "Twelve-Day War," when U.S. forces first hit Fordow, Natanz, and Isfahan before a ceasefire on 24 June 2025 [4][5]. When a fresh round of diplomacy collapsed over the same irreconcilable demand — Washington insisting on zero enrichment, Tehran on its right to enrich — the United States and Israel reached again for the instrument used against Iraq's Osirak reactor in 1981 and Syria's Al-Kibar reactor in 2007: the preventive strike.

By the time a Pakistan-brokered ceasefire took hold on 8 April and the operation was formally declared over on 6 May, the bombing had run for weeks, the regime had been decapitated, and the question that justified the entire enterprise remained unanswered [6]. Did Epic Fury end Iran's path to a weapon, merely delay it — or do what the strike on Osirak is now widely judged to have done to Iraq, converting a hedged, ambiguous, deniable program into a determined, concealed, and unstoppable one?

The analysis that follows argues the strike has almost certainly *not* ended the program. It has bought a delay measured in months, not years, and has materially raised — not lowered — the probability that the Middle East ends this decade with more nuclear-armed states than it began it. That is a forward-looking judgment, not a reported fact; but it rests less on speculation about what we can see inside Iran today than on what half a century of preventive counterproliferation, documented in the scholarly record, tells us happens next.

What the bombs actually destroyed

The physical picture is partial and contested by design. CSIS's post-strike assessment described Iran's enrichment facilities at Fordow and Natanz as "decimated" by the strikes, while noting that Iran still possesses roughly 400 kilograms of 60-percent-enriched uranium whose "exact location remains unknown" [7]. The IAEA, for its part, has not been granted access to verify what portion of that stockpile survived or where it now sits [8]. That is the crux of the entire assessment: centrifuge cascades can be rebuilt, and a stockpile of enriched uranium, if dispersed before the strikes, does not evaporate when a hall collapses — and no inspector can now confirm where it is.

That stockpile is the heart of the problem, and the trajectory was stark well before the first bomb fell. Under the 2015 Joint Comprehensive Plan of Action, Iran's "breakout" time — the interval needed to produce enough weapons-grade uranium for a single device — was held to roughly twelve months [9]. The U.S. withdrawal from the deal in 2018 and Iran's progressive escalation of enrichment collapsed that cushion: as Colin Kahl testified to Congress in February 2023, what "back in 2018 … would have taken about 12 months" had become "about 12 days" [10]. By 2025 Iran's stockpile of 60-percent-enriched uranium had grown into the hundreds of kilograms — CSIS put it at roughly 400 kilograms at the time of the strikes [7] — and independent analysts assessed breakout as a matter of weeks [9][11]. Weapons-grade is 90 percent — but the engineering distance from 60 to 90 percent is far shorter than the climb from natural uranium to 60. The hardest enrichment work was already done.

Fordow compounds the problem. The facility sits under roughly 80 to 90 meters of rock — built, by Iran's own logic, precisely to survive the kind of campaign now being waged against it. Whether the bunker-penetrating munitions reached its cascades is unknown outside a handful of intelligence services, and quite possibly uncertain even to them — and CSIS itself stresses that the exact location of Iran's surviving enriched-uranium stockpile remains unknown [7]. This is the recurring epistemic trap of counterproliferation: the attacker rarely knows what it actually destroyed, and the target has every incentive to ensure it never finds out.

The historical record is not encouraging

The case for preventive strikes rests almost entirely on a single celebrated precedent — Israel's 1981 destruction of Iraq's Osirak reactor — and that precedent does not survive scrutiny.

Conventional memory holds that Osirak set Saddam Hussein's bomb back by years. The archival record assembled over the following four decades says the opposite. Drawing on interviews with Iraqi nuclear scientists, the political scientist Målfrid Braut-Hegghammer found that before the raid Iraq's nuclear effort was directionless and disorganized — exploratory, with no explicit political mandate for a weapon [12][13]. The strike supplied the mandate. It converted a meandering reactor program into a focused, secret, dispersed weapons effort built around clandestine enrichment — concealed so effectively that the 1991 Gulf War coalition badly underestimated its scope until inspectors uncovered it after the war [13][14]. The bombing did not delay the Iraqi bomb. It created the program that pursued it.

Osirak is not an isolated data point. In a systematic study of the category, the political scientist Dan Reiter reached a blunt conclusion: limited strikes against nuclear, biological, and chemical programs have a poor record, with net delays that are minimal even in the cases most often cited as successes, because of weak intelligence and the target's capacity to adapt, conceal, and disperse [15]. The deeper structural point, argued by Scott Sagan among others, is darker still: the act of striking can resolve a target state's internal ambivalence *in favor* of the bomb, by demonstrating precisely the vulnerability a nuclear deterrent exists to cure.

Syria's Al-Kibar reactor, destroyed by Israel in September 2007, is the standing counterexample — a strike that did appear to end a program. But Al-Kibar proves the narrowness of the success, not its generality. It was a single, above-ground, North-Korean-modeled plutonium reactor, not yet operational, in a state with no indigenous nuclear-industrial base and no enrichment fallback [16][17]. Iran in 2026 is the structural inverse: decades of accumulated expertise, a dispersed enrichment infrastructure, a deep stockpile of near-weapons-grade material, and hardened sites built in anticipation of exactly this attack. The conditions that made Al-Kibar a clean kill are the conditions Iran does not have — and never did.

The thing that cannot be bombed

The most important asset Iran possesses is not at Natanz or Fordow. It is in the heads of its scientists and in the centrifuge designs it has held for three decades.

Iran's enrichment program traces to the A.Q. Khan network, the Pakistani-centered black market that sold Iran P-1 centrifuge components and P-2 drawings in the 1990s — the same catalog that armed Libya and North Korea [18][19]. The network was dismantled in 2004; the knowledge it transferred was not. This is the irreducible fact of mature proliferation: enrichment know-how, once acquired and indigenized, cannot be bombed. Buildings can be flattened and rebuilt. A national cadre of nuclear engineers, and the design knowledge they carry, can only be deterred, bought off, or assassinated one by one — and assassination, the Iraqi case suggests, hardens resolve rather than dissolving it.

This is why the more defensible reading is that Epic Fury has produced a delay, not a termination — and a delay whose length depends entirely on variables the bombing cannot touch: how much enriched material survived dispersed and, by the IAEA's own admission, unaccounted for [8]; whether Fordow's cascades are intact [7]; and, above all, what the regime emerging from this war concludes it must do never to be this vulnerable again.

The decision the strike forced

That last variable is decisive, and the historical pattern points the wrong way. Before 28 February, Iran was a textbook nuclear *hedger* — accumulating capability and latency while remaining, formally and ambiguously, a non-weapons state inside the Non-Proliferation Treaty. Hedging is a reversible posture, the stance of a state keeping its options open precisely because it has not yet decided the costs of the bomb are worth paying [20].

Operation Epic Fury demolished the case for ambiguity. The forward-looking analytic judgment here — speculative, but consistent with the documented Iraqi precedent and the preventive-strike literature — is that a surviving Iranian leadership will conclude that latency without a weapon invited annihilation, and that the only durable guarantee against regime decapitation is a deliverable device, built in secret and dispersed beyond the reach of any future air campaign. No Iranian nuclear test has occurred, and one remains unlikely during active hostilities: assembling and testing a device *under sustained bombardment* is extraordinarily difficult, and CSIS notes the program's surviving remnants are degraded though not eliminated [7]. But the war's duration is not its aftermath. The strategic logic the bombing has implanted in any successor regime is the opposite of reassuring: it has taught Tehran that the bomb is not the provocation but the insurance.

The cascade

The consequences do not stop at Iran's borders. Proliferation rarely occurs in isolation: the near-acquisition of a weapon by one regional power generates security anxieties in its neighbors that, historically, only matching capability resolves — the dynamic analysts call a proliferation cascade.

Saudi Arabia is the bellwether. Crown Prince Mohammed bin Salman has said so on the record, twice — to CBS in 2018 ("if Iran developed a nuclear bomb, we will follow suit as soon as possible") and to Fox in 2023 ("if they get one, we have to get one") [21][22]. Riyadh's conduct is classic hedging: pressing for enrichment and fuel-cycle autonomy while remaining formally within the NPT. The Carnegie Endowment's analysis of NPT-withdrawal dynamics holds that Iranian and Saudi proliferation could pull other regional actors toward their own programs, creating a "highly destabilizing" arms race [23] — and the driver is not Iran's program alone but eroding confidence in U.S. security guarantees, precisely the erosion this war has accelerated across the Gulf. Turkey, which already hosts U.S. weapons under NATO nuclear sharing and whose leadership has periodically voiced interest in an independent capability, sits one tier behind; Egypt and the UAE appear in the cascade literature as second-order followers.

The symmetry is grim. A strike justified as nonproliferation may, by destroying the credibility of the security guarantees that kept Gulf states non-nuclear, prove the most powerful proliferation accelerant the region has ever seen.

The precedent for everyone else

The final cost is paid outside the Middle East entirely. Every state weighing the value of a nuclear deterrent has now watched a non-nuclear NPT member — one that negotiated a verified agreement, then saw it abandoned by the other party — subjected to regime-decapitating strikes. The lesson for Pyongyang, and for any future proliferator, is the one North Korea has preached for two decades: states that give up the bomb, or never get it, get bombed; states that test one do not. Libya, which surrendered its Khan-network weapons program in 2003 and whose leader was killed in 2011, is the case study every aspiring proliferator already cites. Epic Fury has just supplied a second.

The verdict

Did Operation Epic Fury end Iran's bomb? On the available evidence and the weight of the historical record, no. It degraded infrastructure that can be rebuilt while leaving intact the two things that actually constitute a nuclear program — a stockpile that may have survived dispersed, and a cadre of expertise that cannot be struck from the air. It has, at most, purchased a delay measured in months.

The harder judgment is the one the Iraqi archives force. The preventive strike does not merely fail to end determined programs; it tends to *create* the determination it was meant to forestall, and to convert hedging neighbors into proliferators in their own right. The analytic conclusion — forward-looking, and offered as analysis rather than reported fact — is that the most likely strategic legacy of Epic Fury is not a denuclearized Iran but an Iran more committed to a weapon than it was on 27 February, in a region with fewer security guarantees and more aspiring nuclear states than at any point since the NPT entered force. The war was launched to answer the nuclear question. On the historical record, its most probable achievement is to have made the answer yes, and the question permanent.

Method note: this is sourced analysis of a real conflict. The 2026 Iran war — Operation Epic Fury, the strikes on Natanz/Fordow/Isfahan, the killing of Ali Khamenei, the IAEA verification blackout, and the April–May ceasefire — is documented in the cited primary and major-outlet reporting; it post-dated some earlier reference cutoffs and is reported here exactly as Ukraine or Sudan would be. Historical and scholarly claims (Osirak, Al-Kibar, the A.Q. Khan network, JCPOA breakout timelines, the MBS statements) are sourced to real, verifiable publications, and the MBS quotes are genuine, precisely attributed pre-2026 remarks. Forward-looking conclusions about whether the strike delayed or guaranteed an Iranian weapon, and about the regional cascade, are labeled as analysis and rest on the cited factual and scholarly record; they are not presented as reported events.

Open in the interactive site →