India and the Global Nuclear Regime

India has spent nearly sixty years arguing with an international order it helped to design and then declined to join. The argument has been remarkably consistent: the objection is to the regime’s discrimination, not to non-proliferation itself. What makes the case worth examining is that India was eventually accommodated without ever conceding the point — admitted to nuclear commerce by waiver rather than by treaty, and inside three of the four export-control regimes it once denounced.

The order India refused

The Treaty on the Non-Proliferation of Nuclear Weapons is the load-bearing instrument of the global nuclear order, and every Indian objection is an objection to one of its structural features rather than to its purpose. Understanding the architecture precisely is what allows the Indian case to be stated as a legal argument rather than a grievance.

The three pillars and the central bargain

• The NPT was opened for signature in 1968, entered into force in 1970, and rests on a bargain among three pillars.
  • Non-proliferation — non-nuclear-weapon states undertake not to acquire nuclear weapons and to accept IAEA safeguards on all their nuclear material.
  • Disarmament — Article VI obliges parties to pursue negotiations in good faith on effective measures relating to cessation of the arms race and to nuclear disarmament.
  • Peaceful uses — Article IV recognises an “inalienable right” to develop nuclear energy for peaceful purposes, with technology transfer as the reward for abstention.
• The treaty has attained near-universal membership, with 191 states parties and four outsiders: India, Pakistan, Israel never joined, and North Korea withdrew in 2003.
• Its historical achievement is real and should be conceded before it is criticised. Kennedy’s forecast of twenty-five nuclear states by the 1970s did not materialise; South Africa dismantled its arsenal, Ukraine, Belarus and Kazakhstan transferred theirs, and Libya abandoned its programme in 2003.

The 1 January 1967 cut-off

• Article IX(3) defines a nuclear-weapon state as one that manufactured and exploded a device before 1 January 1967. This is the hinge of the entire Indian argument.
  • The definition is chronological, not normative — it confers permanent legal privilege on the five states that happened to test first, and it is incapable of being satisfied by anyone else.
  • The five so defined are the United States, the Soviet Union/Russia, the United Kingdom, France and China — the same five holding permanent seats and the veto in the Security Council, which fuses nuclear status with institutional power.
  • No later achievement, restraint or record can change a state’s category. India’s position is that a treaty which makes status depend on a date is not a rule but a settlement.

The asymmetries the treaty builds in

• Safeguards are asymmetric by design. Non-nuclear-weapon states accept comprehensive, full-scope safeguards under INFCIRC/153; the five accept only voluntary offer agreements covering facilities they choose to declare.
• Article VI is an obligation of conduct, not of result — to pursue negotiations, with no deadline, no benchmark and no enforcement.
• The 1995 Review and Extension Conference extended the treaty indefinitely and unconditionally, converting a twenty-five-year experiment into a permanent settlement. For India this was decisive: whatever leverage the non-nuclear states held expired the moment extension was granted without a disarmament timetable attached.

The NPT’s cut-off date does not describe a category of states. It creates one, and closes it.

India inside the negotiation, and out of it

India was not a bystander to the treaty it refused. It participated in the drafting, shaped the criteria by which the draft was to be judged, and withdrew when those criteria were abandoned — a sequence that distinguishes principled refusal from mere abstention.

The ENDC years and Resolution 2028

• India sat in the Eighteen Nation Disarmament Committee (ENDC) through the drafting years, and its delegate V.C. Trivedi, Secretary in the Ministry of External Affairs, led the delegation for most of the negotiation.
• India was a mover of the Eight-Nation Joint Memorandum, whose principles the General Assembly adopted as Resolution 2028 (XX), 19 November 1965. The resolution set five tests for an acceptable treaty.
  • It should contain no loopholes permitting proliferation in any form.
  • It should embody an acceptable balance of mutual responsibilities between nuclear and non-nuclear states.
  • It should be a step towards general disarmament, and particularly nuclear disarmament.
  • It should include workable provisions to ensure its effectiveness.
  • It should not prejudice regional nuclear-weapon-free arrangements.
• The draft satisfied one of the five. It closed proliferation loopholes for the have-nots while providing no balance of obligations and no disarmament step, and India withdrew from the drafting rather than sign what it had helped frame.

Trivedi’s formulation

• Trivedi’s argument distinguished two kinds of discrimination, and it is the distinction rather than the rhetoric that matters. India could live with an asymmetry in weapons; it would not accept one in peaceful and economic development.

“The non-nuclear countries are prepared to tolerate a nuclear weapons apartheid, but not an atomic apartheid in their economic and peaceful development.”

— V.C. Trivedi

• The second Indian formulation was that the treaty pursued not the non-proliferation of nuclear weapons but the non-proliferation of nuclear-weapon powers — a freeze on membership of the club rather than a reduction in the danger.
• The commercial dimension was raised in the same period, with Indian negotiators objecting that the arrangements on peaceful nuclear explosions amounted to an atomic commercial super-monopoly, since the five would sell PNE services that others were forbidden to develop.

The grounds of India’s objection to the NPT

India’s case is best set out as a series of discrete legal and security propositions rather than as a single complaint, because each carries a different weight and each attracts a different counter-argument.

Discrimination as the primary ground

• The treaty divides sovereign equals into two juridical classes and makes the division permanent, which India argues is incompatible with the sovereign equality asserted in Article 2(1) of the UN Charter.
• Jaswant Singh gave the argument its most cited modern statement, describing the arrangement as a sort of international nuclear apartheid and treating the 1995 indefinite extension as the moment at which the legitimation of the five arsenals became irreversible.

“The legitimization of nuclear weapons implicit in the unconditional and indefinite extension of the NPT is also irreversible.”

— Jaswant Singh

• The analogy to apartheid is doing specific work and is not merely polemical: it names a system in which rights depend on a classification the classified cannot exit.

Horizontal proliferation restrained, vertical proliferation unrestrained

• Horizontal proliferation is the spread of weapons to new states; vertical proliferation is the expansion and qualitative improvement of existing arsenals.
• The NPT restricts only the first. There is no ceiling on warhead numbers, no bar on new delivery systems, no restraint on modernisation, and no limit on the accuracy or yield of what the five may build.
• The pattern has held. Global stockpiles fell after 1986 for reasons of superpower politics rather than treaty obligation, and the current direction is upward: China’s stockpile roughly doubled since 2020 to about 620 warheads, and every one of the five is running a modernisation programme.
• India’s demand, made consistently since the 1960s, is a reciprocal bargain — the non-nuclear states forgo acquisition in return for the nuclear states forgoing retention, not merely forgoing transfer.

Article VI and the ICJ’s advisory opinion

• India’s strongest legal argument is that the five are in continuing default on their own treaty obligation, and that a bargain whose consideration has failed cannot bind the other side.
• The International Court of Justice, in its advisory opinion of 8 July 1996 on the Legality of the Threat or Use of Nuclear Weapons, found unanimously that the obligation is stronger than the treaty’s language is usually taken to imply.

“There exists an obligation to pursue in good faith and bring to a conclusion negotiations leading to nuclear disarmament in all its aspects under strict and effective international control.”

— International Court of Justice, 1996

• The words “bring to a conclusion” convert Article VI from an obligation to talk into an obligation to achieve a result. Fifty-six years after entry into force, no such negotiation has been concluded or seriously begun.
• The Court also found that the threat or use of nuclear weapons would generally be contrary to international humanitarian law, dividing only on the extreme case of a threat to state survival. This is the ground on which India’s proposed convention prohibiting use rests.

The permission the treaty leaves open

• Articles I and II forbid the transfer of nuclear weapons and of control over them, but do not forbid deployment on allied territory under the owner’s control.
• The result is that NATO nuclear sharing — American weapons based in several European non-nuclear-weapon states, with their aircraft assigned delivery roles — coexists with the treaty, while an identical arrangement outside an alliance would be treated as proliferation.
• India’s point is not that the arrangement is illegal but that it demonstrates the treaty polices who owns weapons rather than where they sit, which is a weaker discipline than it presents itself as being.

The peaceful nuclear explosion question

• Peaceful nuclear explosions (PNEs) were, in the 1960s, a serious technological proposition — excavation, canal and harbour construction, gas stimulation, seismic experiment — pursued by both superpowers.
• The physics defeats the legal distinction: a peaceful explosion and a weapons test are the same event, differing only in declared purpose.
• The Indian objection was to the one-sidedness rather than to the restriction. Article V offered non-nuclear states the benefits of explosions conducted by the five, while forbidding them to conduct their own — dependence written into the treaty.
• The bilateral Peaceful Nuclear Explosions Treaty, 1976 confirmed the asymmetry: the superpowers regulated their own PNEs by agreement while the regime denied the option to others.
• The category is now closed, since the CTBT bans all nuclear explosions without regard to declared purpose — which retrospectively vindicates the Indian argument that the distinction was never technically real.

The safeguards asymmetry

• Non-nuclear-weapon states accept comprehensive safeguards over all nuclear material in all activities, with design review, material accountancy, reporting and inspection.
• The five accept only voluntary offer safeguards, selecting the facilities they submit; in practice the IAEA inspects a token fraction of them.

The security ground

• Accession is possible only as a non-nuclear-weapon state. The treaty has no category India could join without first disarming, and no amendment creating one has ever been tabled with any prospect of adoption.
• Signing would therefore require India to surrender the arsenal while facing two nuclear-armed neighbours, one of which is the state whose 1964 test triggered India’s reconsideration and with which the boundary remains undelineated.
• The security assurances offered to non-nuclear states — UNSC Resolution 255 (1968) and later Resolution 984 (1995) — promise only that the Council will act, which is worth nothing when a permanent member is the adversary. This is the specific failure India experienced in 1964–65.
• India’s conduct, on its own account, supplies the treaty’s object without its signature: no transfers, a voluntary testing moratorium, comprehensive export controls and an India-specific safeguards agreement.

NPT provision	What it does	India’s objection
Article IX(3) — 1 January 1967 cut-off	Defines nuclear-weapon status by date of first test	Confers permanent privilege by chronology; the category cannot be entered
Articles I and II	Bar transfer of weapons and of control	Permit deployment on allied territory; police ownership, not location
Article III — safeguards	Comprehensive for the many, voluntary for the five	Intrusion applied only downwards
Article IV — peaceful uses	Recognises an inalienable right	Right conditioned on accession, and denied through supplier cartels
Article V — PNEs	Benefits of explosions conducted by the five	Dependence written into the text; the distinction is not physically real
Article VI — disarmament	Obligation to pursue negotiations in good faith	Unfulfilled for five decades; the ICJ read it as an obligation to conclude
1995 extension	Indefinite and unconditional	Removed the only leverage the non-nuclear states possessed

Testing the objection: the case against India’s position

An answer that only states India’s grounds has not examined them. The counter-case is serious, is made by scholars sympathetic to India as well as hostile to it, and has to be met rather than mentioned.

• Refusal weakens the regime by demonstrating that persistence is rewarded. India held out for forty years and was then granted access to nuclear commerce; the lesson available to any determined state is that the costs of defiance are temporary.
• It supplies a precedent that others cite. Pakistan’s case for equal treatment in the NSG rests almost entirely on the Indian precedent, and China has used that symmetry to block India — so the precedent has already cost India something concrete.
• It sits awkwardly with India’s own advocacy of a rules-based order. India argues for multilateralism, for the primacy of treaty law, and for Security Council reform in the name of legitimacy, while standing outside one of the system’s central rules.
• It left India outside the NSG, which is the practical residue of the objection. The 2008 waiver delivered the trade; membership would have delivered the vote, and that remains withheld.
• A realist reading dissolves the moral framing entirely. On this account India’s opposition was never principally about discrimination — it was the rational conduct of a state preserving a capability option in an anarchic system, with the language of equity supplying the diplomacy that made the option affordable.
• The counter to the realist reading is that India’s conduct outran what interest alone would explain: it accepted four decades of technology denial, forwent testing between 1974 and 1998, and continued tabling disarmament proposals that a purely instrumental state would have dropped.

The measured conclusion is that India’s position is internally coherent and has been vindicated in practice, since the international system found a way to accommodate it without the treaty. What the position cannot resolve is the tension it creates — arguing for rules from outside one of them — and India has managed that tension by over-performing on conduct rather than by answering the objection.

If India signed both treaties: the implications

The question is a counterfactual and is best answered by separating what would change legally, what would change materially, and what would change in India’s diplomatic position. The three answers point in different directions.

Signing the NPT

• India could accede only as a non-nuclear-weapon state, which means the arsenal would have to be dismantled and all fissile material placed under comprehensive safeguards. There is no accession route that preserves the weapons.
• Deterrence would end against two nuclear-armed neighbours, and India would depend for its security on assurances the Security Council has never been able to deliver against a permanent member.
• The strategic programme’s remaining facilities — the fast breeder, the naval reactor programme, the reprocessing plants kept off the civilian list — would come under safeguards, ending the autonomy for which the 2006 separation was designed.
• The gains would be modest. Access to nuclear commerce India already has by waiver; the NSG seat would presumably follow; and India would gain standing in the disarmament debate it currently argues from outside.
• The net judgment is that accession would convert an accommodated de facto nuclear power into a disarmed treaty party in exchange for benefits it has largely already obtained — which is why no Indian government of any party has contemplated it.

Signing the CTBT

• The material cost is lower, because India has observed a voluntary moratorium since May 1998 and has already conducted the tests its designs required.
• The cost that remains is the loss of the option. India’s thermonuclear yield from 1998 has been publicly disputed by some of its own scientists, and signature would foreclose validation of a boosted or thermonuclear design.
• Signature would not bind others. With the treaty not in force and Russia having de-ratified in November 2023, India would accept a legal constraint in a field where the constraint is not yet general.
• The diplomatic gain would be substantial — it would remove the last formal argument against India’s non-proliferation credentials, and would strengthen the NSG bid.
• The Indian position has converged on the substance without the signature: a de facto test ban offered unilaterally, with the legal option retained, which is the outcome the 1996 objection was designed to preserve.

India has already conceded the CTBT’s substance and withheld only its signature; it can concede nothing of the NPT’s substance without ceasing to be a nuclear-armed state.

India and the Comprehensive Test Ban Treaty

The CTBT episode is often collapsed into the NPT one, and the two questions are different. India refused the NPT because it was discriminatory; it refused the CTBT because it was ineffective and, in its final form, coercive. Keeping the two objections distinct is what makes India’s position legible as a policy rather than as a habit of refusal.

The negotiation India joined and then blocked

• India was an early and enthusiastic sponsor of a comprehensive test ban, having called for one since 1954, and it entered the Conference on Disarmament (CD) negotiations from 1994 as a supporter.
• Two issues bogged the negotiation down: the scope of “comprehensive”, and the verification arrangements.
  • On scope, the treaty banned nuclear explosions but left untouched sub-critical experiments, hydrodynamic testing and computer simulation, which are precisely the techniques by which established arsenals are maintained and refined without explosions.
  • On verification, states divided over whether an international monitoring system would amount to an intrusion on sovereignty, and over how on-site inspections would be triggered.
• India’s ambassador to the CD, Arundhati Ghose, declared in June 1996 that the emerging text was not conceived as a step towards disarmament and was not in India’s security interest, and blocked consensus on forwarding the draft.
• The treaty was then taken out of the CD and adopted by the UN General Assembly, September 1996 — a procedural manoeuvre India described as circumventing the consensus rule of the negotiating body.

The first objection: the treaty is not comprehensive

• A ban on explosions freezes capability at existing levels rather than reducing it, and it freezes them unevenly.
  • States with hundreds of past tests and validated designs lose nothing; states with a handful lose the ability to develop further.
  • Laboratory and sub-critical methods, which only the advanced programmes possess, allow qualitative improvement to continue lawfully.
• India’s demand was accordingly a linkage: it would accept the test ban if the five accepted a time-bound framework for elimination, so that the freeze became the first step of a process rather than the last.

“India cannot accept any restraints on its capability if other countries remain unwilling to accept the obligation to eliminate their nuclear weapons.”

— India’s statement to the Conference on Disarmament, June 1996

• Jaswant Singh’s later summary of the objection was that the treaty was “neither comprehensive nor related to disarmament” — two failures rather than one.

The second objection: Article XIV was coercive

• Article XIV makes entry into force conditional on ratification by the 44 states listed in Annex 2 — those possessing research or power reactors at the time of negotiation.
• India was named in Annex 2, which meant the treaty could not enter into force without India’s ratification, and India read this as an attempt to impose an obligation on a state that had refused consent — a clause it called attempted duress, since consent to be bound is the foundation of treaty obligation.
• This is why the two objections are analytically distinct. The NPT objection is about equity; the CTBT objection is about efficacy and consent. A non-discriminatory test ban would still have failed India’s first test, and a treaty without Article XIV would still have failed it too.

	Objection to the NPT	Objection to the CTBT
Core ground	Discrimination — two permanent classes of state	Ineffectiveness — freezes capability without reducing it
Legal form	Sovereign inequality; unfulfilled Article VI	Consent — Article XIV names India and coerces ratification
What would cure it	A disarmament timetable binding on the five	Linkage to time-bound elimination; deletion of Article XIV
India’s conduct	Refusal, and a parallel record of non-transfer	Voluntary moratorium since 1998, a de facto ban offered without signature
Current position	No prospect of accession without disarming	Substance conceded; signature withheld

From refusal to unilateral moratorium

• After Pokhran II in May 1998, India declared a voluntary unilateral moratorium on further explosive testing and offered to convert it into a de facto test ban, while retaining the sovereign right to resume.
• The moratorium became a settled element of the 2003 doctrine and was reaffirmed in the commitments underpinning the 2005–08 accommodation, so it now carries diplomatic weight that a signature would only formalise.

The treaty’s position in 2026

• The CTBT has 188 signatories and 179 ratifications, but it has not entered into force and there is no prospect that it will.
• Of the 44 Annex 2 states, 35 have ratified, leaving nine outstanding: China, Egypt, Iran, Israel, Russia and the United States have signed but not ratified; India, Pakistan and North Korea have not signed.
• Russia de-ratified in November 2023, explicitly to mirror the American position, though it remains a signatory and therefore bound not to defeat the treaty’s object and purpose.
• The testing norm has come under strain.
  • An American directive of 30 October 2025 on resuming nuclear testing was clarified days later as covering system and non-critical tests.
  • Putin’s order of 5 November 2025 sought proposals on resuming tests. No explosive test has followed, and the taboo has held in practice.

The FMCT and the Conference on Disarmament

• A Fissile Material Cut-off Treaty (FMCT) would ban the production of highly enriched uranium and plutonium for weapons, capping arsenals at the material already held.
• India supports it, and has done so consistently — the formula India uses is a treaty that is universal, non-discriminatory and internationally verifiable, the same three adjectives it applies to every disarmament instrument it endorses.
• The Shannon Mandate of 1995, named for the Canadian ambassador who negotiated it, established the CD’s mandate to negotiate such a treaty and deliberately left open the question of scope, so that both sides could enter negotiations without conceding it.

Stocks versus future production

• The unresolved question decides everything the treaty would do. A cut-off covering future production only freezes existing disparities; one covering existing stocks would require reductions and would function as a disarmament measure.
• The five prefer a future-production treaty because their stockpiles are large and, in most cases, production has already stopped — they would concede nothing.
• Pakistan demands that stocks be included, because its fissile material holdings are smaller than India’s and a future-production freeze would lock in the asymmetry permanently.
• India’s position is closer to the five’s, since a stocks treaty would constrain its own material for a growing arsenal — which is the honest way to read India’s support and should be stated rather than glossed.

The Conference on Disarmament and the consensus rule

• The CD in Geneva is the single multilateral disarmament negotiating forum, with 65 members, and it operates by consensus on everything including its own agenda.
• Pakistan has blocked the commencement of FMCT negotiations since 2009, and the consensus rule means one state’s objection halts the body entirely. The CD has not concluded a treaty since the CTBT text of 1996.
• The deadlock is instructive about the wider order: the machinery is not broken so much as accurately reflecting the absence of agreement, and moving the negotiation elsewhere, as was done with the CTBT and the TPNW, produces instruments the relevant states do not join.

Why supporting the FMCT and refusing the NPT is not a contradiction

• The FMCT would apply the same obligation to every state, including the five, which satisfies the non-discrimination test that the NPT fails.
• It would be verified internationally on the same terms for all, rather than comprehensively for some and voluntarily for others.
• It caps the vertical dimension that the NPT leaves untouched, which is the specific gap India has complained of since the 1960s.
• India’s consistency is therefore in the criteria rather than in the answers: it supports instruments that bind all states equally and refuses those that entrench a division, and the FMCT and the NPT fall on opposite sides of that line.

The export-control regimes

The treaty regime restrains states that join it; the export-control regimes restrain everyone by controlling supply. For India these were the operative instruments of denial for thirty-four years, and its relationship with them is now the most revealing measure of where it sits in the order.

The Nuclear Suppliers Group

• The NSG was created in 1974–75 as a direct response to India’s Pokhran I explosion, which had used plutonium from a Canadian-supplied research reactor moderated with American heavy water, both provided for peaceful purposes.
  • Originally the London Club of seven suppliers, it went beyond the Zangger Committee’s trigger list by covering states outside the NPT and adding conditions the treaty did not require.
• Its instrument is a set of guidelines applied through national law, operating in two parts.
  • Part 1 covers the trigger list — reactors, reprocessing and enrichment plant, heavy water, source and special fissionable material.
  • Part 2, adopted at Warsaw in 1992, covers dual-use goods and technology with civil applications and possible weapons relevance.
  • A full-scope safeguards condition was added in 1992, requiring recipients to have comprehensive safeguards on all their nuclear activities — a condition India by definition could not meet.
• The group has 48 participating governments, takes all decisions by consensus, and has no treaty basis, no secretariat of its own and no enforcement mechanism beyond national implementation.
• Its effect on India was decisive and is the reason the civil programme stagnated: not a ban on India’s weapons, but a fuel and technology embargo on India’s power reactors.

The 2008 waiver and the membership bid

• The clean waiver of 6 September 2008 exempted India from the full-scope safeguards condition, permitting members to trade with a state outside the NPT — the only such exemption ever granted.
• India applied for full membership in May 2016, and the bid came to a head at the Seoul plenary of June 2016, where it failed.
• China’s objection was procedural in form and substantive in effect. Beijing argued that admission of any non-NPT state required a non-discriminatory, criteria-based approach applicable to all, and pressed a two-step process: first agree criteria for non-NPT applicants generally, then consider individual applications.
  • Pakistan filed its own application days after India’s, which converted the question into a paired one and gave China a symmetry argument it has used ever since.
  • Several other members raised concerns about admitting states outside the NPT without a general rule, so the obstacle is not China alone, though China is the one that has never moved.
• Rafael Grossi, then chairing the NSG process and later Director General of the IAEA, produced a criteria paper intended to square the circle.
  • It set out conditions any non-NPT applicant could satisfy — a separation plan, safeguards, a testing moratorium and support for the FMCT.
  • It was circulated in December 2016 but never adopted — resisted both by members who thought the criteria too weak and by China’s insistence on agreeing a general rule first — and no membership decision has been taken since.
• India’s counter-argument has three parts.
  • The NSG is an export-control body, not a non-proliferation treaty organisation, so NPT membership is not logically a condition of participation.
  • India’s record on export control is stronger than that of several members, and it already harmonises its national list with NSG guidelines.
  • France was a founding member of the NSG in 1974 and acceded to the NPT only in 1992 — eighteen years inside the group as a non-party — which disposes of the claim that membership has always required accession.

What membership would actually add

• The honest answer is: less than the campaign suggests, but not nothing.
  • Trade is already permitted by the waiver, so membership would not unlock supplies India cannot now obtain.
  • What it would add is a vote and a veto — the right to shape guidelines rather than to receive them, and protection against a future tightening that narrows the waiver.
  • It would also secure the waiver institutionally. A waiver granted by consensus can in principle be revisited by consensus; membership converts an exemption into a seat.

Inside three regimes, outside the fourth

• India was admitted to the Missile Technology Control Regime in June 2016, to the Wassenaar Arrangement in December 2017, and to the Australia Group in January 2018.
• The pattern is instructive. In each of the three, China is not a member, so there was no veto to overcome; in the NSG, China is.

Regime	Concerned with	Founded	India’s status
Nuclear Suppliers Group	Nuclear and nuclear dual-use exports	1974–75	Outside; waiver 2008, membership blocked
Missile Technology Control Regime	Missile and UAV delivery systems	1987	Member, June 2016
Wassenaar Arrangement	Conventional arms and dual-use goods	1996	Member, December 2017
Australia Group	Chemical and biological precursors	1985	Member, January 2018

India’s own export-control architecture

• The Weapons of Mass Destruction and their Delivery Systems (Prohibition of Unlawful Activities) Act, 2005 criminalises the manufacture, acquisition, transfer and brokering of nuclear, chemical and biological weapons and their delivery systems, and was enacted specifically to implement UN Security Council Resolution 1540 (2004).
  • The 2022 amendment closed the financing gap, prohibiting the financing of any activity banned under the Act and empowering the government to freeze, seize and attach related funds.
• The SCOMET list — Special Chemicals, Organisms, Materials, Equipment and Technologies — is India’s national control list, administered through the foreign trade policy and harmonised with NSG, MTCR, Wassenaar and Australia Group guidelines.
• The record of non-transfer is the substantive point: no Indian nuclear technology has ever been found in a proliferation network, which contrasts sharply with the record of at least one NPT party in the region.

India was excluded from the supply regimes for conduct it did not repeat, and admitted to three of them for conduct it never abandoned.

India’s disarmament diplomacy

India’s disarmament record is usually treated as decoration on a nuclear-weapons story. It is better read as the other half of the same argument: the claim that the objection is to discrimination rather than to non-proliferation is only credible because India has tabled more disarmament proposals than any other state possessing nuclear weapons.

The 1954 standstill and the road to the PTBT

• Nehru’s proposal of 2 April 1954 for a standstill agreement on nuclear testing, made in Parliament after the American Castle Bravo test at Bikini, was the first such call by any state and predates every test-ban instrument.
  • Its reasoning was framed in terms of the effects of radioactive fallout on populations remote from the testing, which anticipated the humanitarian argument by six decades.
• India carried the proposal into the UN through the 1950s and was among the original signatories of the Partial Test Ban Treaty of 1963, which banned tests in the atmosphere, in outer space and under water.
• The PTBT is the pivot of India’s later CTBT argument: it drove testing underground without slowing it, which is exactly the pattern India predicted the CTBT would repeat with simulation.

The 1978 proposal and the Six-Nation Initiative

• In 1978 India proposed at the General Assembly the negotiation of an international convention prohibiting the use of nuclear weapons — an attempt to attack the legitimacy of the weapons rather than their numbers.
  • The proposal has been tabled annually since, and it is the resolution most closely connected to the ICJ’s 1996 finding on humanitarian law.
• In 1982 India called for a nuclear freeze — a halt to production of fissile material for weapons, to warhead production and to delivery-system development.
• The Six-Nation Five-Continent Peace Initiative, launched on 22 May 1984, brought together Argentina, Greece, India, Mexico, Sweden and Tanzania — the leaders being Alfonsín, Papandreou, Indira Gandhi, de la Madrid, Palme and Nyerere.
  • Its founding statement demanded an immediate halt to testing, production and deployment of nuclear weapons.
  • Its practical achievements were modest but real: it contributed to the atmosphere in which Gorbachev announced a unilateral testing moratorium in August 1985, and it opened scientific verification contacts between Soviet and American institutions.
  • The group, sometimes called the Delhi Six, wound down around 1989 as superpower arms control took over the agenda.

The Rajiv Gandhi Action Plan, 1988

• Presented to the Third Special Session on Disarmament (SSOD-III) in June 1988, the Action Plan for a Nuclear-Weapon-Free and Non-Violent World Order remains the only such plan ever tabled by a head of government.
• Its design, rather than its rhetoric, is what makes it examinable.
  • A time-bound framework — elimination of all nuclear weapons in three phases by 2010, running roughly six years, six years and ten years.
  • Binding commitments at each stage, so that obligations were legal rather than declaratory and each phase’s completion conditioned the next.
  • Verification built into every phase, with treaties and institutions to insure against non-compliance rather than verification deferred to the end.
  • Universality and non-discrimination — every state possessing or capable of possessing weapons undertakes the same obligations, which is the direct inversion of the NPT’s structure.
  • Collateral measures — parallel reduction of conventional forces and prohibition of the weaponisation of outer space, on the reasoning that states will not give up nuclear weapons while conventional imbalances remain.
  • A negotiated ban on use as an early confidence-building step, and the eventual restructuring of the international order around non-violence.
• Why it failed is a more useful question than its content.
  • It was tabled at the end of the Cold War’s most dangerous decade, when the superpowers had just concluded the INF Treaty and preferred bilateral arms control they controlled to a multilateral framework they did not.
  • The five had no incentive to accept binding elimination while their security doctrines were built on the weapons.
  • India’s own credibility was contested even then, since Rajiv Gandhi had authorised weaponisation in the same period, and the plan’s failure is often cited as the moment India concluded that advocacy alone would not work.
• The plan retains standing because its principles — universality, non-discrimination, verifiability, time-bound phases — are the criteria India still applies to every instrument it is asked to join.

The standing resolutions and the CD papers

• India has pressed its case through the Non-Aligned Movement as well as the General Assembly, and NAM’s standing demand for a nuclear weapons convention is substantially of Indian authorship.
• India tables three resolutions annually in the General Assembly’s First Committee, and they are the operational form of its declared position.
  • Convention on the Prohibition of the Use of Nuclear Weapons — a legally binding prohibition on use and threat of use, pending elimination.
  • Reducing Nuclear Danger — measures against unintentional, accidental and unauthorised use, including de-alerting, lowering operational readiness and reviewing doctrines that permit launch on warning.
  • Measures to Prevent Terrorists from Acquiring WMD, which is the one that passes without a division and reflects a genuine convergence.
• India tabled a Working Paper on Nuclear Disarmament in 2006 and has submitted successive papers to the CD proposing a Nuclear Weapons Convention, and it remains the only nuclear-armed state calling unambiguously for such a convention.
• The 2003 doctrine embeds the position: continued commitment to a nuclear-weapon-free world through global, verifiable and non-discriminatory disarmament, alongside participation in FMCT negotiations and observance of the testing moratorium.

The TPNW and why India stays out

• The Treaty on the Prohibition of Nuclear Weapons was adopted in 2017 and entered into force on 22 January 2021, prohibiting development, testing, possession, stationing, use and threat of use outright.
• It has 95 signatories and about 74 states parties, and its First Review Conference is set for 30 November to 4 December 2026 in New York. No nuclear-armed state and no NATO member has joined.
• India did not participate and does not support the treaty, and its stated grounds are procedural and substantive rather than hostile to the objective.
  • The negotiation took place outside the Conference on Disarmament, which India regards as the only forum where an instrument binding the nuclear-armed states can be negotiated.
  • The treaty does not provide for verifiable, universal and non-discriminatory disarmament, having no verification machinery capable of confirming elimination and no participation by the states that would have to disarm.
  • India’s view is that it creates no new customary law and does not bind non-parties.
  • Nina Tannenwald’s account of a nuclear taboo supplies the treaty’s best defence — that stigmatisation is itself a disarmament instrument, working on the legitimacy of the weapons rather than on their numbers.
• The awkwardness is worth naming. The TPNW is universal and non-discriminatory in exactly the way India demands, and India’s objection rests on forum and verification rather than on equity — which is a narrower ground than the one it uses against the NPT.

The critique: declaratory since 1998

• The standing criticism is that India’s disarmament diplomacy became declaratory after Pokhran II — resolutions tabled annually and voted down annually, at no cost and to no effect.
• The supporting evidence is that India expanded its arsenal through the same period in which it tabled elimination proposals; its stockpile of about 190 warheads now exceeds Pakistan’s, and the triad has been completed.
• The defence is that no nuclear-armed state has done more, that the proposals are legally serious rather than symbolic, and that a state cannot be asked to disarm unilaterally as the price of being believed.
• The fairest reading is that India’s disarmament advocacy is sincere as a preference and inert as a policy: it describes the world India would prefer, and it does not describe anything India will do first.

Why disarmament does not happen

Every disarmament proposal since 1946 has failed for reasons that are structural rather than accidental, and setting them out is the difference between describing the failure and explaining it.

The security dilemma and the universality requirement

• Disarmament is a collective action problem with a first-mover penalty: the state that disarms first is worse off than every other, so no state moves first and the equilibrium holds.
• Universality is a requirement, not an aspiration. Elimination that leaves one holdout is worse than no elimination at all, since the holdout acquires a monopoly — which is why India insists that any instrument bind all states.

Verification and intrusiveness

• Verification of elimination is qualitatively harder than verification of limits. Counting deployed missiles is feasible; proving the absence of a small number of concealed warheads or of undeclared fissile material is close to impossible.
• The intrusiveness required — access to military sites, weapons laboratories, industrial facilities and personnel — is incompatible with what states will accept as sovereignty, and the more complete the disarmament the more intrusive the regime must be.

The irreversibility of knowledge

• Weapons can be dismantled; the knowledge cannot. Every state that has once weaponised retains the designs, the trained personnel and the industrial base to rebuild.
• Disarmament therefore produces virtual arsenals rather than a world without weapons, and the strategic problem shifts to the speed of reconstitution — which some analysts regard as more unstable than open possession.

Alliances, doctrine and domestic constituencies

• Extended deterrence multiplies the number of states with an interest in retention. Some thirty states shelter under American, Russian or allied nuclear guarantees, and each would have to be reassured before its protector could disarm.
• Military-industrial and bureaucratic constituencies entrench the weapons institutionally. Laboratories, delivery-system programmes and service branches acquire a stake in retention, which is the organisational argument Scott Sagan applies to the whole nuclear enterprise.
• Modernisation locks in the future. All nine possessors are running modernisation programmes; a warhead or submarine entering service now has a service life stretching to mid-century, which converts a present decision into a fifty-year commitment.

The collapse of the bilateral architecture

• The instruments that actually reduced arsenals were bilateral and are almost all gone. The ABM Treaty lapsed on American withdrawal in 2002; the INF Treaty ended in 2019; the Open Skies Treaty unravelled from 2020.
• New START expired on 5 February 2026, with no successor in force. For the first time since 1972, nothing legally binding constrains the two largest arsenals, and the verification and inspection regime that came with the treaty has gone with it.
• The multilateral machinery has stalled in parallel: the CD has produced nothing since 1996, the CTBT is not in force, and the 11th NPT Review Conference closed without a consensus document in May 2026 — the third consecutive failure.
• The emerging competition is also trilateral rather than bilateral, with China’s expansion making the arithmetic of parity intractable, and it extends into missile defence, hypersonic systems, anti-satellite weapons, cyber and autonomous systems, none of which is covered by any arms-control instrument.

The socio-economic cost of the arms race

• World military expenditure reached $2,887 billion in 2025, an eleventh consecutive real-terms increase, equal to 2.5% of world GDP — the highest share since 2009.
• The concentration is extreme: the United States spent $954 billion, China about $336 billion, Russia $190 billion, and India $92.1 billion, the fifth largest, up 8.9% on the year.
• The guns-versus-butter argument is about opportunity cost rather than waste as such. Every unit of expenditure on deterrence is a unit not spent on health, education, nutrition or infrastructure, and the trade-off falls hardest on states whose social deficits are largest.
  • The human security literature reframes the whole calculation, arguing that for most people the operative threats are disease, hunger, displacement and environmental degradation, and that state-centric security spending misallocates against them.
• Development economics adds a second cost. Arms imports absorb foreign exchange without productive return; defence research absorbs scarce scientific manpower; and for developing states the debt incurred is not offset by any growth effect.
• The Soviet case is the standard illustration — its economic collapse was substantially aggravated by the share of output devoted to matching American military spending, which is the strongest historical evidence that arms races can be self-defeating.
• The counter-argument should be stated. Deterrence is cheaper than the war it prevents; defence spending generates employment and spin-off technology; and the Cold War arms race, on the realist reading, purchased forty years without great-power war.
• The victims’ side of the ledger is usually omitted: radiation exposure at test sites and among downwinders, the environmental legacy of production complexes, and the unresolved problem of waste — costs borne by populations that were never consulted.
• The deeper cost is not the money but the choice it conceals: an arms race is a standing decision about which insecurities a state treats as urgent, taken repeatedly and almost never deliberated.

The accommodation of 2005–08

Between 2005 and 2008 the international system did something it had never done: it accommodated a nuclear-armed state outside the NPT without requiring it to join, disarm or abandon its position. The mechanism was a chain of instruments, each conditional on the last, and each with its own bargain.

The July 2005 joint statement

• On 18 July 2005, Prime Minister Manmohan Singh and President George W. Bush issued a joint statement in Washington in which the United States undertook to seek full civil nuclear cooperation with India and to adjust American law and international regimes accordingly.
• India’s reciprocal commitments were specific: separate civil from military facilities, place the civil facilities under IAEA safeguards, sign an Additional Protocol, continue the testing moratorium, work for an FMCT, and maintain comprehensive export controls harmonised with NSG and MTCR guidelines.
• The statement’s opening formulation — that the United States would work to achieve full civil nuclear cooperation with India as a “responsible state with advanced nuclear technology” — was the diplomatic invention that made everything else possible, since it created a category outside the NPT’s binary.

The Separation Plan, March 2006

• Announced during Bush’s Delhi visit in March 2006, the plan identified which facilities would be declared civilian and therefore safeguarded, and it was negotiated in India rather than dictated.
• Fourteen of India’s twenty-two thermal power reactors were placed on the civilian list in phases to 2014; the rest, along with the strategic programme, stayed out.
• What India kept out is the substance of the bargain: the Prototype Fast Breeder Reactor and the fast breeder programme, the CIRUS research reactor and the Dhruva reactor, the naval propulsion programme, and the reprocessing capacity supporting them.

The Hyde Act, December 2006

• The Henry J. Hyde United States–India Peaceful Atomic Energy Cooperation Act amended American law to exempt India from the full-scope safeguards requirement of the US Atomic Energy Act, and was signed in December 2006.
• It attached conditions that Indian critics read as extra-territorial policy prescription.
  • Cooperation terminates if India tests, and the President must seek the return of transferred materials and equipment.
  • Annual presidential certification to Congress that India is complying with its non-proliferation commitments.
  • A reporting requirement on whether India is working with the United States on Iran, which was read in India as conditioning nuclear cooperation on alignment in West Asia.
• The Indian government’s position was that the Hyde Act is American domestic law binding on the American executive, and that India’s obligations run only from the bilateral agreement — a distinction the Left rejected as legally naive.

The 123 Agreement, 2007

• The bilateral Agreement for Cooperation Concerning Peaceful Uses of Nuclear Energy, concluded in July–August 2007, is named for Section 123 of the US Atomic Energy Act, which requires such an agreement before nuclear commerce. India extracted three things Washington had initially refused.
  • Fuel supply assurances, including support for building a strategic reserve sufficient to run safeguarded reactors for their lifetime, and a commitment to convene friendly suppliers if supplies were disrupted.
  • Advance consent to reprocess US-obligated spent fuel in a new national facility under IAEA safeguards — the arrangements were negotiated separately and concluded in 2010.
  • No explicit testing prohibition in the text, with the right of termination on a year’s notice and a consultation requirement — leaving India’s testing option legally intact even if politically costly.

The India-specific safeguards agreement, 2008

• Approved by the IAEA Board of Governors, 1 August 2008, the agreement is India-specific rather than the standard INFCIRC/153 comprehensive form, applying to declared civilian facilities only.
• Its most contested feature is the relationship between safeguards in perpetuity and corrective measures. The preamble records India’s understanding that it may take corrective measures to keep its reactors running if fuel supply is disrupted — a conditional exit on India’s reading, a loophole on the non-proliferation community’s.

The NSG waiver, September 2008

• After two special plenaries and sustained American lobbying, the NSG adopted a clean, unconditional waiver on 6 September 2008, exempting India from the full-scope safeguards condition.
• Clean meant no explicit testing condition and no periodic review in the text — resisted by Austria, Ireland, New Zealand, the Netherlands, Switzerland and others, and secured partly through an Indian statement reaffirming the moratorium.
• The effect was that India became the only nuclear-armed state outside the NPT permitted to conduct nuclear commerce, and cooperation agreements followed with Russia, France, the United Kingdom, Canada, South Korea, Kazakhstan, Argentina, Mongolia, Namibia, Australia and Japan among others.

The domestic politics

• The deal was contested harder in India than in the United States. The Left parties withdrew support from the UPA government in July 2008 on the ground that it compromised strategic autonomy and tied India to American foreign policy.
• The government survived the resulting confidence vote of 22 July 2008, which is the point at which the deal became irreversible domestically. The related question of Parliament’s role in shaping the 123 Agreement belongs with the study of India–US relations, where the accountability argument rather than the nuclear regime is the subject.

What the deal actually was

• C. Raja Mohan reads it as a geopolitical rather than commercial instrument — not a reactor sale but the recognition of India as a major power, the de facto legitimation of its weapons, and its induction as a responsible participant in international politics.
• Rajeswari Pillai Rajagopalan treats it as a direct consequence of American recognition of India’s major-power status and of its strong non-proliferation record, which locates the causation in India’s conduct rather than in American generosity alone.
• Sumit Ganguly locates the American motive in balancing a rising China, on the reasoning that a stronger partnership with Delhi served Washington’s Asian strategy.
• S. Jaishankar assessed it as a turning point in India’s foreign policy, raising India’s global standing and opening opportunities well beyond the nuclear field.

	What India conceded	What India obtained
Facilities	14 of 22 power reactors to the civilian list, safeguarded in perpetuity	Fast breeder, CIRUS, Dhruva, naval programme kept outside
Testing	Reaffirmed the voluntary moratorium	No treaty-level test ban; the legal option retained
Safeguards	India-specific agreement and an Additional Protocol	Corrective measures clause tied to fuel-supply assurance
Fuel cycle	Accepted safeguards on imported fuel	Reprocessing consent and a strategic fuel reserve
Regime	Harmonised export controls with NSG and MTCR lists	Clean NSG waiver; entry to three regimes by 2018

The critique from both sides

• From the left, the deal compromised strategic autonomy by embedding India in an American framework, subjected the civil programme to conditions set by a foreign legislature, and constrained fissile material production for weapons by placing reactors under safeguards.
• From the right, Bharat Karnad argued that the arrangement effectively capped the arsenal at a level chosen by others and foreclosed testing in practice, since resumption would collapse the whole structure — trading a permanent strategic constraint for a temporary energy benefit.
• The common ground between the two critiques is that India accepted real limits in exchange for benefits that were slow to materialise, and the liability stand-off that followed made that criticism look prescient for fifteen years.

The civil nuclear programme and what the deal delivered

The three-stage design

• Homi Bhabha’s three-stage plan was a response to a resource endowment: modest uranium — identified domestic resources of a few hundred thousand tonnes, and of high extraction cost — against about 846,000 tonnes of thorium, on official Indian estimates roughly a quarter of the world total.
  • Stage one — pressurised heavy water reactors burning natural uranium, moderated by heavy water, producing power and plutonium-239 in spent fuel.
  • Stage two — fast breeder reactors fuelled by that plutonium, breeding more fissile material than they consume from a uranium-238 blanket, and progressively adding thorium.
  • Stage three — thorium-based reactors converting thorium-232 into fissile uranium-233, which is the point at which India’s reserves become a long-term energy base.

Where each stage actually stands

• Stage one is mature. The PHWR fleet is indigenous, the 700 MWe design is in series construction, and India builds and operates these reactors without foreign assistance.
• Stage two has just begun. The 500 MWe Prototype Fast Breeder Reactor at Kalpakkam, operated by BHAVINI, attained first criticality on 6 April 2026 — more than fifteen years behind its original schedule and at more than twice its sanctioned cost, at about ₹8,181 crore.
  • Pre-project work has been approved for two further 500 MWe fast breeder units at the same site, and the associated reprocessing facility is now expected around 2029.
• Stage three remains distant. The Advanced Heavy Water Reactor (AHWR), a 300 MWe thorium-fuelled design, has been at the proposal stage for two decades and is not under construction, and the Kamini research reactor is the only U-233-fuelled reactor India has operated.
• The realistic conclusion is that thorium is a strategic asset, not a near-term fuel, and that India’s capacity growth to 2047 will come overwhelmingly from stage one and from imported light water reactors.

The uranium famine

• Between 1974 and 2008 the NSG embargo bit not on weapons but on fuel for power reactors, and the effect was measurable in operations rather than in headlines.
• Reactors ran at load factors as low as 50% in the mid-2000s for want of fuel, and capacity additions slowed while domestic mining struggled with low-grade ore and local opposition.
• Post-waiver imports reversed this within a few years, and safeguarded reactors running on imported uranium returned to rated capacity — the clearest single measure of what 2008 delivered.

The fleet in 2026 and the targets

• India operates about 24 reactors with roughly 8.78 GW installed, all of them until 2025 run by the state operator, the Nuclear Power Corporation of India Limited (NPCIL), supplying about 3% of electricity generation — a small share by any comparison, and the fact that most disputes about Indian nuclear policy tend to ignore.
• Eight reactors are under construction, and sanctioned projects take capacity to roughly 22 GW by 2031–32.
• The declared target is 100 GW by 2047, split 54 GW from the state operator — the existing and under-construction 22 GW plus 32 GW of indigenous PHWRs and light water reactors — and 46 GW from central and state enterprises, state governments, private companies and joint ventures.
• A Nuclear Energy Mission was announced in the Union Budget 2025-26, with about ₹20,000 crore committed to indigenous small modular reactor development.

Small modular reactors

• Three indigenous SMR designs are running in parallel.
  • BSMR-200, a 220 MWe design, the most advanced, with Tarapur in Maharashtra as the lead site.
  • SMR-55, a 55 MWe unit aimed at remote locations, islands and smaller industrial clusters.
  • A high-temperature gas-cooled reactor of up to 5 MWth aimed at hydrogen production, being developed at a BARC facility in Andhra Pradesh.
• At least five indigenous SMRs are targeted by 2033, and the attraction is factory fabrication, shorter build times and siting flexibility rather than any breakthrough in physics.

Fuel security

• India’s identified domestic uranium is a small fraction of world reserves and of high extraction cost, which is why the embargo bit on power generation rather than on weapons. Supply arrangements now run with Kazakhstan, Russia, France, Uzbekistan, Canada and Australia, among others.
• Movement has been rapid. Cameco of Canada contracted to supply nearly 22 million pounds of uranium concentrate between 2027 and 2035; an administrative arrangement with Australia in July 2026 finally operationalised the 2014 civil nuclear agreement; and a long-term arrangement with Uzbekistan advanced in August 2026.
• Annual requirement is around 1,900 tonnes and will rise more than tenfold on the 2047 trajectory, so supplier diversification is now a security question rather than a commercial one.

The case for nuclear power, and against

• For: nuclear is dispatchable baseload that solar and wind are not; it has among the lowest lifecycle emissions of any source; it is extraordinarily land-efficient compared with utility-scale renewables in a land-scarce country; and it is difficult to see a credible path to India’s net-zero-by-2070 commitment that leaves it out.
• Against: capital cost and construction time are the chronic weakness, and the PFBR’s fifteen-year overrun is India’s own evidence.
  • Siting and land acquisition have generated sustained local resistance at Kudankulam, Jaitapur and Kovvada.
  • Spent fuel and high-level waste have no permanent disposal solution anywhere, and the accident tail-risk is small in probability and enormous in consequence.
• The reasonable position is that nuclear is necessary but not sufficient, and that on any realistic trajectory it remains a single-digit share of Indian generation into the 2040s.

Non-power applications

• Medicine — radioisotope production for diagnosis and therapy, teletherapy units, and radiopharmaceuticals supplied through a national network of cancer centres.
• Agriculture and food — mutation breeding has produced dozens of released crop varieties, and radiation processing extends shelf life and meets phytosanitary requirements for export.
• Water — the desalination plant at Kalpakkam using reactor waste heat, and technologies for removing arsenic, fluoride and uranium from groundwater.
• Industry and research — non-destructive testing, sterilisation of medical products, isotope hydrology and accelerator-based research.
• These are the practical content of the NPT’s Article IV promise, and the reason technology denial was experienced in India as a development question and not only a strategic one.

Liability: the fifteen-year stand-off and its resolution

The international principle and India’s departure from it

• The international liability conventions — Paris (1960), Vienna (1963) and the Convention on Supplementary Compensation (1997) — rest on exclusive channelling of liability to the operator.
  • The purpose is practical: it avoids litigating causation through a supply chain of thousands of components, and it requires only one entity in the chain to insure.
  • Liability is strict and no-fault, capped in amount and limited in time, with the state providing a second tier above the cap.
• India departed from this in 2010, and the reason was Bhopal. A catastrophic industrial accident in which a foreign supplier’s design and a domestic operator’s practices were both implicated, followed by a settlement widely regarded as inadequate, made exclusive operator liability politically impossible to legislate.

The Civil Liability for Nuclear Damage Act, 2010

• The Act imposed strict, no-fault liability on the operator, capped at ₹1,500 crore, with the government liable above that up to the rupee equivalent of 300 million SDRs.
• Two provisions froze the market for fifteen years.
  • Section 17(b) gave the operator a right of recourse against the supplier where the incident resulted from an act of the supplier or its employee, including equipment or material with patent or latent defects or sub-standard services.
  • Section 46 provided that the Act was in addition to and not in derogation of other law, leaving tort and criminal liability open against operators and, by extension, suppliers.
• The suppliers’ objections were consistent and were not unreasonable.
  • India was the only country whose law allowed suppliers to be pursued as a matter of statute.
  • Supplier exposure was uncapped, since no ceiling was fixed for them as it was for the operator.
  • The insurance requirement was unquantifiable, because the exposure was.
  • The absence of a definition of nuclear damage left the scope of claims open.
  • Criminal liability could be pursued against operator and supplier alike.
• The government’s 2015 workarounds addressed the politics and not the law.
  • An official FAQ read section 17(b) as permissive rather than mandatory.
  • An India Nuclear Insurance Pool of ₹1,500 crore was created to cover operator and supplier risk.
  • Suppliers were not persuaded, because a government interpretation does not bind a future litigant or court.
• India ratified the Convention on Supplementary Compensation in 2016, aligning with the international architecture without amending the provisions that conflicted with it.

The consequences

• Jaitapur, planned as the largest nuclear project in the world at six EPR units with a French supplier, has been stalled since its 2009 memorandum.
• Kovvada in Andhra Pradesh, intended for American reactors, has not progressed beyond discussion.
• Kudankulam remained the only foreign presence in the Indian fleet precisely because its agreements predate the 2010 Act.
• The judgment that follows is uncomfortable: the 2008 accommodation delivered fuel and standing but almost no reactors, and the obstacle was Indian law rather than the international regime.

The SHANTI Act, 2025

• The Sustainable Harnessing and Advancement of Nuclear Energy for Transforming India Act, 2025 was introduced on 15 December 2025, passed both Houses by 18 December, and received assent on 20 December 2025.
• It repealed both the Atomic Energy Act, 1962 and the Civil Liability for Nuclear Damage Act, 2010, replacing India’s entire nuclear legal architecture in a single instrument.
• Liability was restructured, not abolished. The flat operator cap gave way to a graded cap keyed to reactor capacity, alongside the international ceiling of 300 million SDRs per incident, with the government covering the excess.
• The right of recourse was narrowed decisively. It now runs only where the contract expressly provides for it, or where the incident results from an act done with intent to cause damage — the statutory recourse for patent or latent defects and sub-standard services is gone.
• No equivalent of section 46 was carried forward, so the channelling of liability to the operator moves closer to the international norm and the statute no longer expressly preserves the parallel route through tort and criminal law.
  • Whether that route is actually closed is unsettled: the Supreme Court has asked whether anything in the Act could preclude a constitutional court from awarding fair and just compensation.
• Generation was opened to private companies and joint ventures, ending the state monopoly, while companies incorporated outside India are barred from holding a licence — so foreign capital must enter through domestically incorporated entities.
• Sensitive activities — enrichment, reprocessing, spent fuel management and heavy water production — remain reserved to government entities, and the Atomic Energy Regulatory Board received statutory recognition for the first time, having existed since 1983 only by executive order.

	CLND Act, 2010	SHANTI Act, 2025
Operator cap	Flat ₹1,500 crore	Graded, ₹100 crore to ₹3,000 crore by capacity
Recourse against suppliers	Statutory, for patent or latent defects and sub-standard services	Only as the contract provides, or for acts done with intent
Other laws	Section 46 left tort and criminal liability open	Not carried forward — channelling closer to the international norm
Who may operate	State monopoly through the public operator	Private companies and joint ventures admitted
Foreign entities	Not applicable	Barred from licences; investment via Indian-incorporated entities
Regulator	AERB by executive order	AERB on a statutory footing

The case for the reform and the case against

• For: the 2010 regime demonstrably did not work, producing no reactors and no compensation architecture worth the name.
  • The graded cap raises the ceiling for large reactors, and statutory backing for the regulator is a real gain in independence.
  • Private participation is arithmetically necessary if 46 GW of the 2047 target is to come from outside the public operator.
• Against: liability was narrowed at the moment private operators entered, which reverses the usual sequence in which the entry of profit-seeking operators tightens rather than loosens accountability.
  • Caps of any size are grossly inadequate against a major accident — Fukushima’s costs are estimated in the hundreds of billions of dollars.
  • Removing statutory recourse weakens the supplier’s incentive to build safely, since the cost of a defect now falls on the operator and ultimately the state.
  • Closing the parallel route through tort sits uneasily with Indian constitutional doctrine on absolute liability for hazardous activity and with the Article 21 jurisprudence on remedy, which is the ground on which the Act has been challenged before the Supreme Court, with the challenge pending.
  • Regulatory independence remains contested in substance, since the board’s members are appointed by the same government that owns the operators it regulates.

India spent fifteen years insisting that suppliers must answer for defects, and then removed that right in the same statute that let private operators in.

Where India stands in the order in 2026

• India is a de facto nuclear-armed state accommodated by waiver, not by treaty, with about 190 warheads, a completed triad and a stockpile that now exceeds Pakistan’s.
• It is inside the MTCR, Wassenaar and the Australia Group, and outside the NSG, which is the single unfinished item of the 2008 settlement.
• The regime it refused is under greater strain than at any time since its creation.
  • The 11th NPT Review Conference ended without a consensus document in May 2026 — a third consecutive failure across sixteen years.
  • New START has expired and the CTBT remains out of force with Russia de-ratified; the testing norm was strained by American and Russian statements in late 2025, though no explosive test has followed.
• India’s own conduct has not changed through any of this: the moratorium holds, the export controls hold, the record of non-transfer holds, and the resolutions are tabled every autumn.
• The domestic settlement has changed decisively, with the SHANTI Act removing the obstacle that had made the 2008 accommodation commercially inert, and with the fast breeder finally critical after four decades of promise.

India has obtained almost everything the treaty would have given it and conceded almost nothing the treaty would have required, which is a diplomatic achievement without precedent. What it has not obtained is an answer to the objection its own case invites — that a state arguing for a rules-based order from outside one of that order’s central rules is relying on an exception it would deny to others. India manages that contradiction by over-performing on conduct; management is not resolution.