WATCHTOWER · INVESTIGATION
Israel has nuclear weapons. The Federation of American Scientists and Stockholm International Peace Research Institute identify it as nuclear-armed and estimate about 90 warheads. Israel has not officially acknowledged possession; the exact inventory remains unverified. FAS’s current 2026 assessment; SIPRI Yearbook 2026 summary, chapter 8.
The secrecy deserves scrutiny. So do the program’s foreign assistance, clandestine procurement and the allegations involving the United States. There are proven export violations and documented espionage. There is also a serious unresolved uranium-diversion case. They are not interchangeable proof that every claim about stolen American nuclear secrets is true.
Our position, developed in the clearly labeled analysis below, is that Israel should have no nuclear weapons and should enter a credible, internationally verified disarmament process. That judgment rests on the catastrophic nature of the weapons, the danger of unaccountable power, and the record of civilian harm and nuclear rhetoric examined here. It applies a standard we would also demand of every other nuclear-armed government.
Reported and examined by DaddyBruce95 · November 7, 2026. This investigation uses current arms research, declassified U.S. records, legal documents, official statements and attributed reporting. Estimates, intelligence assessments, criminal convictions, unresolved allegations and editorial conclusions are identified separately.
1. What is known—and what “about 90” means
SIPRI’s 2026 assessment estimates approximately 90 Israeli warheads as of January 2026. FAS’s current 2026 table gives the same approximate stockpile. These organizations’ research networks overlap; matching numbers should not be advertised as two wholly independent discoveries. Neither number is an inspection certificate or a count released by Israel. SIPRI’s dated estimate; FAS table and uncertainty notes.
A stockpile estimate is also different from the amount of material that analysts believe a country could potentially turn into weapons. Production potential is not a verified inventory. Likewise, a reserve or nondeployed classification in a research table does not establish that weapons are unusable or that there is no capacity to prepare them for use.
The important distinction is between a strong conclusion supported by decades of evidence and the many details that remain hidden. A report can state that Israel is nuclear-armed without pretending to know every warhead, operational order or target. This article makes the first claim. It does not invent the rest.
2. Opacity was a political arrangement, not an absence of concern
Declassified U.S. diplomacy shows how much officials understood and how carefully they handled public acknowledgment. In a July 19, 1969 memorandum to President Richard Nixon, Henry Kissinger distinguished stopping Israeli possession from, at minimum, keeping it from becoming an established international fact. The memo records disagreement inside the U.S. government over leverage, assurances and military supplies. Read the original memorandum in the State Department’s historical series.
An October 8, 1969 memorandum examined Israel’s proposed assurances and the ambiguity of saying it would not become a nuclear power or be the first to introduce nuclear weapons. The documents show officials disputing what those formulations meant. They must not be converted into a reliable no-first-use pledge or a promise that Israel possessed nothing. Kissinger’s analysis of the proposed assurances.
These records matter because secrecy can be negotiated and politically useful. It is not evidence that nobody knew, nobody suspected, or no government had choices. The archive reveals decisions about what to press, what to accept and what to say publicly. Those decisions belong in an accountability story alongside the weapons themselves.
3. The program was built with foreign assistance
The historical record does not support a simple story of one country independently producing an arsenal in isolation. Archival research by Avner Cohen and William Burr documents France’s central role in the Dimona project and the concealed activities that U.S. visits in the 1960s failed to uncover. Their work also distinguishes what visiting officials believed from what later evidence showed. Historical study and declassified inspection records.
A December 2, 1960 U.S. Joint Atomic Energy Intelligence Committee statement identified a French-Israeli reactor project and associated activity interpreted as weapons-related. That is contemporaneous intelligence evidence of a developing program, not proof that a finished arsenal already existed in 1960. Some contemporary forecasts were uncertain or inaccurate. Original 1960 intelligence statement.
Norway’s heavy-water assistance, involving a British route, is another part of the history. A June 1959 U.S. record of discussions with a Norwegian official describes peaceful-use conditions and inspection rights. The assistance was real, but saying it came with no conditions whatsoever would erase part of the documentary record. Archive’s examination of the Norway–UK–Israel connection; Original June 1959 memorandum.
Foreign assistance, safeguards, deception and subsequent weapons development are connected questions. They are not all the same act, nor does every nuclear-cooperation agreement prove that its participants knowingly supplied a completed bomb. The history is strongest when the particular transaction and its conditions remain visible.
4. Vanunu’s disclosures—and the limits of claims about testing
In October 1986, former Dimona technician Mordechai Vanunu’s disclosures to the Sunday Times supplied inside information about the program. He was captured by Israeli intelligence, brought to Israel and sentenced to 18 years in prison; he was released in 2004. Those disclosures became an important part of the public evidentiary record. They are not a measurement of the stockpile four decades later. Archival history and the 1986 disclosures.
The 1979 Vela satellite incident remains disputed evidence of a possible nuclear test and possible Israeli involvement. SIPRI’s 2026 assessment says there is no conclusive open-source evidence that Israel has tested its nuclear weapons. Neither a conclusively proved test nor a categorical denial is established here. The disputed Vela record; SIPRI’s current qualification.
Opacity is a reason to demand evidence and transparency. It is not permission to transform every unresolved episode into a certainty. The case for accountability does not depend on settling a disputed test through assertion.
5. NUMEC: a serious U.S. uranium-diversion suspicion
The Nuclear Materials and Equipment Corporation, or NUMEC, operated a nuclear-materials plant in Apollo, Pennsylvania. Significant accounting losses became the subject of federal investigation and suspicions that uranium had been diverted to Israel. The central distinction is between material unaccounted for and material proved stolen. The former is documented; the latter was not conclusively established by the public investigations reviewed here. GAO’s December 1978 report, digest and conclusions.
A March 9, 1972 CIA memorandum treated diversion by NUMEC officials as a “distinct possibility.” That is a serious intelligence assessment. Its material-accounting categories do not turn automatically into a quantity delivered to Israeli weapons production. Original CIA memorandum.
An August 26, 1977 CIA record of a briefing for Representative Morris Udall said the agency had not obtained hard intelligence clearly linking NUMEC to later Israeli uranium-based weapons production. The difference between that statement and other officers’ suspicions is part of the case, not a reason to suppress either record. Original CIA briefing record.
This is not a choice between declaring a solved theft and pretending nothing serious happened. An unresolved diversion allegation involving nuclear material deserves scrutiny precisely because the stakes are so high. It still requires the verbs that the evidence supports: suspected, assessed, investigated and not conclusively determined.
6. The U.S. investigations had failures of their own
GAO’s December 18, 1978 report, Nuclear Diversion in the U.S.? 13 Years of Contradiction and Confusion, could not determine whether diversion occurred. It described federal investigations as limited, uncoordinated and less than adequate, and said restrictions on access to CIA and FBI documentation prevented full corroboration. That is an institutional accountability finding even without a solved diversion case. Read the declassified GAO report.
In a February 1979 follow-up, GAO continued to dispute the denial of information but explained its inability to compel disclosure. Restricted access does not itself prove the suspected theft. It does explain why a review cannot honestly be advertised as having answered everything. GAO’s follow-up on access to records.
A November 2, 2009 Nuclear Regulatory Commission letter to Senator Arlen Specter later said a records review had found no specific evidence establishing diversion. It also could not unequivocally exclude diversion or say that all previously unaccounted-for material had been recovered. The linked document is a scanned NRC letter hosted by an advocacy organization; the agency’s words, rather than the host’s interpretation, are the evidence used here. NRC letter, accession ML092720878.
The responsible conclusion remains demanding: do not invent proof, and do not turn an incomplete record into an acquittal of every institutional failure. Both standards can be maintained at once.
7. Richard Kelly Smyth: a proven export-control case
The Richard Kelly Smyth case is different because a criminal conviction is established. A State Department notice in the Federal Register lists his conviction under the Arms Export Control Act. Contemporary reporting of his December 28, 2001 guilty plea describes one export-control count and one false-Customs-statement count involving a shipment of 50 krytrons to Israel in 1982. Official debarment notice; Contemporaneous report of the plea and prosecutors’ account.
Krytrons are controlled electronic components with nuclear-weapons relevance as well as other military and civilian applications. Their unlawful export is not the same as exporting a complete nuclear weapon, fissile material or American bomb blueprints. Nor does the conviction, by itself, establish the final use of each component. Israel disputed the nuclear end-use described in allegations. The plea report and the stated end-use dispute.
Larger shipment totals appeared in the original indictment and later retellings. They should not be presented as though the two-count plea established every allegation in the indictment. A proved export violation is already a substantial finding. It does not need to be enlarged beyond the court record to matter.
8. Arnon Milchan: what the public admission establishes
In a 2013 Uvda interview, film producer Arnon Milchan publicly described clandestine work connected with Israeli security procurement. Contemporary reporting identified activity for LAKAM, the Bureau of Scientific Relations, and described his account of obtaining material for Israel’s nuclear program. The original broadcaster’s page identifies the report as first broadcast in November 2013, despite its later reposting date. Original broadcaster’s page, in Hebrew; Contemporaneous account of the interview.
The admission is relevant. It should also be described for what it is: a participant’s public account, supported and contextualized by reporting, not a U.S. judgment convicting Milchan of every alleged procurement or theft. No such U.S. conviction of him was established in the records reviewed for this article.
His entertainment career does not turn the conduct of one named person into an accusation against an entire industry or against Jewish people. The accountability questions concern individuals, government organizations, transactions and decisions. Broadening the accusation beyond that evidence would weaken the investigation.
9. Pollard’s espionage was real. The nuclear claim needs precision
Jonathan Pollard, a civilian U.S. Navy intelligence analyst, supplied classified material to Israel. The FBI’s case summary records his arrest in November 1985, guilty plea in 1986 and life sentence in 1987. These are established events, not a speculative theory of Israeli collection against the United States. FBI historical case summary.
The CIA’s October 30, 1987 damage assessment describes Israeli collection priorities that included U.S. intelligence about Arab and Pakistani nuclear programs, alongside other military information. Stealing American intelligence about another country’s nuclear activity is not the same as stealing American nuclear-bomb designs. The released assessment does not establish the latter claim. CIA damage assessment, especially its discussion of collection priorities.
That distinction does not make the espionage harmless. Classified intelligence and collection capabilities were compromised. It does keep the chronology and evidence straight: Pollard’s 1980s espionage is not an explanation for the origins of an Israeli weapons capability assessed much earlier. Redacted passages cannot responsibly be filled with whatever a writer suspects.
10. Outside the NPT does not mean outside all law
Israel is not a party to the Nuclear Non-Proliferation Treaty. The IAEA Director General’s July 27, 2026 report records that it remains the regional exception to NPT membership and comprehensive safeguards. That is a treaty-status finding, not an IAEA count of Israeli warheads. GOV/2026/48–GC(70)/13, paragraphs 4–5.
The NPT’s nonproliferation duties distinguish between its parties and define nuclear-weapon-State status using the treaty’s historical cutoff. Israel’s possession does not make it a sixth treaty-recognized nuclear-weapon State. But it is also inaccurate to describe its possession simply as a breach of obligations it accepted under NPT Article II, because it did not join. NPT text, Articles II, III, VI and IX.
Nor is “no safeguards” an accurate description of every nuclear activity in Israel. Limited, item-specific arrangements exist; comprehensive safeguards over all relevant nuclear material are a different system. The IAEA’s own records document a limited agreement involving Israel and the United States. Limited inspection is not an audit of the weapons program. IAEA report explaining the limited arrangement; Current IAEA country record.
Israel is also outside the Treaty on the Prohibition of Nuclear Weapons. Its treaty prohibition cannot simply be described as a promise Israel signed and broke. None of this removes the relevance of the UN Charter, humanitarian law or other applicable obligations. Legal precision does not supply moral approval. UN depositary record.
11. The international calls for disarmament are explicit
The UN General Assembly’s December 1, 2025 resolution on nuclear proliferation in the Middle East specifically called for Israel to join the NPT, place its nuclear facilities under comprehensive safeguards and renounce possession of nuclear weapons. Those are institutional demands, not a criminal court verdict or an inspection-derived inventory. A/RES/80/67, paragraphs 5–6.
On September 17, 2026, the IAEA General Conference adopted a regional safeguards resolution calling for NPT accession, full-scope safeguards and a mutually and effectively verifiable nuclear-weapon-free zone. It addresses all regional states. A resolution does not, by being passed, create a functioning zone or dismantle an arsenal. GC(70)/RES/10.
The IAEA Director General’s 2026 report also records the underlying disagreement: Israel links safeguards to broader stable regional security and arms-control arrangements; other regional states argue that safeguards and a weapons-free zone should help create security rather than wait for a completed peace settlement. These are competing diplomatic positions, not a technical reason that inspection is impossible. The Director General’s account of regional positions.
A September 2026 EU statement likewise supported arrangements freely arrived at by regional states. There is an existing diplomatic framework to pursue. The absence of progress should not be mistaken for the absence of a proposal. EU statement on regional safeguards and the weapons-free-zone process.
12. The nuclear rhetoric cannot be waved away
On November 5, 2023, Heritage Minister Amichai Eliyahu answered a radio question about nuclear bombing of Gaza by describing it as one option. Prime Minister Benjamin Netanyahu rejected the remarks and announced his suspension from cabinet meetings. Eliyahu later characterized the remark as metaphorical. He was not a member of the war cabinet or the security cabinet, and the statement did not establish an authorized nuclear plan. Contemporaneous Israeli reporting of the remark and the government’s response.
Both parts of that record matter. Omitting the repudiation would misrepresent official policy. Omitting the original remark would erase the conduct that required repudiation. A minister’s willingness to discuss nuclear destruction of a populated territory is relevant to scrutiny of political judgment even when it is not an authenticated launch order.
The question is not whether a public remark reveals the exact command structure of a secret arsenal. It is whether leaders treat the lives of civilians with the seriousness that nuclear responsibility demands. The subsequent metaphor explanation does not make the original exchange disappear.
13. Conduct toward civilians belongs in this assessment
In September 2025, the UN Independent International Commission of Inquiry concluded that Israel had committed genocide against Palestinians in Gaza. Its report examined prohibited acts and intent and identified failures of prevention and punishment. That is the Commission’s investigative and legal finding; it is not a final genocide judgment by the International Court of Justice. The Commission’s report A/HRC/60/CRP.3; Official statement of its findings.
That record is relevant to the argument about entrusting extraordinary destructive power to state institutions. It does not prove that Israel used a nuclear weapon in Gaza, that every Israeli shares criminal intent, or that a particular nuclear strike is imminent. None of those claims is established here.
The publication’s position is that documented large-scale harm and findings of grave international crimes should intensify scrutiny of power, secrecy and safeguards. A nuclear arsenal should not be insulated from that scrutiny because conventional operations and nuclear policy occupy different bureaucratic boxes. For the detailed evidentiary record, read our investigation of the September 2025 Commission report.
14. What nuclear use would mean for human beings
The International Red Cross and Red Crescent Movement has warned that adequate national or international assistance capacity does not exist for the aftermath of nuclear use, and that even a limited exchange could have catastrophic, lasting effects on health, the environment, food production and social life. This is a humanitarian assessment grounded in the consequences of the weapons, not a prediction that a specific government will launch one tomorrow. The Movement’s appeal for humanity.
The ICRC’s legal and policy position emphasizes the extreme difficulty of reconciling nuclear use with humanitarian protection. The relevant standard is the effect on human beings, including people who never chose a war. It cannot be reduced to whether one government describes its intentions as defensive. The ICRC’s legal and policy assessment.
The ICJ’s July 8, 1996 advisory opinion said threat or use would generally be contrary to humanitarian law, while leaving unresolved an extreme self-defense circumstance involving the very survival of a state. That unresolved question was not affirmative permission. The Court also affirmed the obligation to pursue and conclude good-faith disarmament negotiations under strict and effective international control. The advisory opinion; UN publication of the Registry’s summary.
A public debate should not hide the Court’s qualification, nor use it to erase the basic rules protecting civilians. The people endangered by nuclear escalation do not become less human because the state invoking survival is an ally.
15. Washington’s silence is a current oversight question
A May 4, 2026 congressional letter to Secretary of State Marco Rubio, publicized by Representative Joaquin Castro’s office on May 5, demanded greater transparency about Israel’s nuclear capabilities and consistent nonproliferation standards. Its signatories argued that Congress needed information about the regional nuclear balance and escalation risks. The signed letter; Official release and list of signatories.
The letter is an oversight demand, not a declassified stockpile assessment. Its importance is political and democratic: elected representatives were challenging official silence. A government can be asked to explain a policy without the questioner pretending to possess its classified files.
The policy argument here is that scrutiny should follow risk and evidence rather than alliance. Applying consistent standards does not require falsely claiming that states with different treaty commitments have committed identical treaty violations. It requires explaining those differences openly while refusing to turn them into permanent immunity from accountability.
16. The evidence does not need exaggeration
Established in the public record: authoritative arms researchers identify Israel as nuclear-armed; historical foreign assistance and U.S. concern are documented; Smyth’s export-control conviction and Pollard’s espionage are real; Milchan publicly described clandestine procurement activity; Israel remains outside the NPT and comprehensive safeguards; and international bodies have called for nuclear transparency and disarmament.
Estimated, disputed or not established here: the exact current inventory and readiness; a conclusive account of the 1979 test controversy; a proved quantity of NUMEC uranium delivered into Israeli weapons; or the blanket claim that Pollard stole American bomb designs. These limits are not an excuse to dismiss the record. They identify what the record can and cannot carry.
The strongest report does not turn allegations into convictions or official silence into absence of evidence. It shows why each claim is believed, what type of source supports it and where further disclosure is needed. That is how an investigation remains useful after the first wave of outrage.
17. Analysis: Israel should have no nuclear weapons
The following is Watchtower’s ethical and policy analysis, distinguished from the documentary findings above. Our conclusion is direct: Israel should have none. A secret nuclear arsenal is not made acceptable by long possession, technical sophistication or the support of powerful allies. Nuclear weapons hold human beings at risk on a scale that no patriotic slogan can make decent.
The combination of secrecy, grave civilian-harm findings and rhetoric entertaining nuclear destruction strengthens the case for change. A government should not be asked only whether its weapons can deter an enemy. It should be asked what safeguards exist against disastrous judgment, who can scrutinize its conduct, and why civilians must live beneath a threat they cannot meaningfully control.
The usual defense is deterrence: weapons are retained to prevent an existential attack. That security concern must be taken seriously, including the safety of Israeli civilians. It does not settle the argument. Deterrence depends on assumptions about decisions, communication and restraint that can fail. Calling a capability defensive does not guarantee that escalation, miscalculation or misuse can be contained.
Nor would we answer Israeli disarmament by cheering nuclear proliferation elsewhere. Iran should not acquire nuclear weapons. Every existing nuclear-armed government should move toward verified elimination. The standard is consistent because the civilians living under all these governments possess the same human worth. Nobody’s children should be collateral in a theory of strategic superiority.
A serious path forward requires more than demanding an instant announcement. It requires honest acknowledgment and reporting, enforceable commitments, internationally credible verification, comprehensive safeguards, and a negotiated process for dismantling the weapons and preventing their return. Regional arrangements must protect all populations and bind participants through reciprocal obligations. Verification is the difference between disarmament and a press release.
Israel’s outside-the-NPT status should therefore be treated as a problem to resolve, not a permanent privilege. The international resolutions offer a direction: renunciation, safeguards and a verifiable regional zone. Progress will require political pressure and diplomacy, including from governments that have spent decades making exceptions for allies.
The demand is for verified disarmament, not attacks on civilian populations or reckless assaults on nuclear facilities. The goal is to remove the capacity and the incentives for catastrophe while protecting people. A campaign for life cannot become indifferent to the lives endangered by its proposed means.
No government should have the power to erase a city and expect gratitude for not using it. God’s name is not permission to destroy a people. A flag is not an exemption from responsibility. Power this destructive demands more scrutiny, not less—and the final objective should be its removal.
18. What peaceful public pressure can demand
Ask representatives for public explanations of nuclear policy, meaningful oversight and consistent nonproliferation standards. Demand disclosure of historical records where lawful, preservation of evidence, and honest distinctions between established espionage, unresolved diversion and current risk. Challenge nuclear threats regardless of who makes them.
Support efforts toward verifiable regional disarmament and the protection of civilians. The relevant work is diplomacy, monitoring, accountability and sustained public attention. Visit Take Action for peaceful ways to engage. Read Our Standards for this publication’s evidence and correction process. Do not threaten, harass or assign collective guilt in the name of disarmament.
Source record and scope
This report is a synthesis of public evidence, not an inspection of a nuclear arsenal, a new forensic examination of NUMEC material or a private intelligence disclosure. It draws on FAS and SIPRI, original U.S. intelligence and diplomatic records, GAO and NRC documents, official legal records, IAEA reports, UN resolutions, the ICJ opinion, attributed interviews and reporting. An archive can host allegations without adopting them; a scanned agency letter is distinguished from the views of its host.
The inventory estimate is dated and qualified. Historical records are not presented as current readiness assessments. Convictions are distinguished from indictment allegations and participant admissions. The IAEA’s circulation of a state’s letter is not treated as the agency’s endorsement of its claims. No technical construction instructions, private details or unverified launch plans are included. Research cutoff and publication date: November 7, 2026.