The Nuclear Ghost: Has Iran’s Atomic Program Really Been ‘Obliterated’?

A claim repeated insistently by the White House — that the June 2025 strikes on three of Iran’s key nuclear facilities “obliterated” the country’s nuclear program — is now being tested by the global nonproliferation community. As Operation Epic Fury entered its second day, striking additional military and governmental targets across Iran, the emerging consensus among technical experts is considerably more cautious than the administration’s language suggests.
The June 2025 strikes — conducted as part of Israel’s 12-day military campaign against Iran, with the United States targeting three specific nuclear sites — did inflict significant damage on Iran’s enrichment infrastructure. Satellite imagery analyzed by CSIS showed substantial destruction at the Natanz enrichment facility, the Fordow fuel enrichment plant, and the Arak heavy water reactor — Iran’s three most significant known nuclear installations.
But “significant damage” and “obliterated” are not synonyms in nonproliferation vocabulary. Nuclear programs are not single facilities — they are distributed knowledge networks, material stockpiles, human expertise, and redundant infrastructure spread across a country’s territory. The scientists who built Iran’s enrichment cascades are still alive. The engineering knowledge encoded in their training did not vaporize when the centrifuges were destroyed. Uranium enrichment technology is, at its core, industrial engineering applied to a specific material challenge — and Iran has been mastering that challenge for thirty years.
The CSIS analysis published on the day of Operation Epic Fury — “Operation Epic Fury and the Remnants of Iran’s Nuclear Program” — acknowledges the critical distinction between destroying known facilities and eliminating a latent nuclear capability. Iran has, at various points over the past decade, enriched uranium to 60% purity — well beyond the 20% threshold for research reactors, and approaching the 90% enrichment level needed for weapons-grade material. Whether stockpiles of highly enriched uranium survived both strike campaigns is a question that only a thorough on-the-ground inspection — the kind that requires a functioning, cooperative government in Tehran — can definitively answer.
This creates a profound strategic paradox. The military operations designed in part to eliminate Iran’s nuclear threat may have simultaneously eliminated the government structures through which a verifiable denuclearization agreement could have been reached. The final round of nuclear negotiations in Geneva, which collapsed without progress just days before Operation Epic Fury, represented the last opportunity for a diplomatic resolution. With Khamenei dead and Iran’s provisional leadership focused on immediate survival and retaliation, the prospect of a negotiated, inspected, verifiable dismantlement of Iran’s nuclear program has receded dramatically.
What the nuclear nonproliferation community fears most in a scenario like this is not the known threats but the unknown ones. Did Iran have undeclared enrichment sites that Western intelligence missed? What happened to the scientists and engineers from Iran’s nuclear program in the chaos of the strikes? Could they resurface elsewhere, carrying knowledge and potentially material? The history of weapons programs in destabilizing states — from the Soviet dissolution to post-2003 Iraq — is littered with cautionary tales about the consequences of destroying state structures without controlling what those structures contained.
There is also the deterrence calculus for every other state watching these events unfold. North Korea, which has developed nuclear weapons and has not been attacked, will draw one set of conclusions. Iran, which remained short of weapons-grade capability and has been struck twice in nine months, provides the counterfactual. The implicit lesson — that nuclear weapons are the ultimate deterrent against regime change — may be the most durable and dangerous legacy of Operation Epic Fury.