On 22 January 2026, in a UNCITRAL Working Group III session (WGIII) on procedural reform of investor-state dispute settlement, India opposed a draft provision giving non-disputing treaty parties the right to make submissions in ISDS proceedings, reasoning from its WTO experience rather than any stated concern about ISDS itself. The position appears difficult to reconcile with India’s own record in the forum, where it has for years pushed for reform more sweeping than most other states would support. A government advocating extensive structural reform nevertheless opposed a narrower procedural measure, for reasons unrelated to the architecture of ISDS itself.
That moment illustrates a larger pattern not yet read as a single question. Since 2024, India has signed agreements reflecting markedly different approaches to whether foreign investors should enjoy direct remedies against the state. With the EFTA states, under the Trade and Economic Partnership Agreement in force from 1 October 2025, India excluded investor protection and investor-state arbitration entirely. The investment chapter covers only promotion and facilitation, and the treaty’s general dispute settlement mechanism does not even apply to it. Oman and the UK, under agreements signed in 2025, follow the same architecture: states may bring claims against each other, investors may not. The India-Brazil BIT, done at New Delhi on 25 January 2020 and in force since 21 December 2025, goes further still, and does so by substitution rather than subtraction: drawing on Brazil’s own cooperation-and-facilitation model, it replaces investor arbitration with a joint committee and a national ombudsman in each country, a dispute prevention procedure that runs ahead of any arbitration, and an arbitral mechanism, once available, confined to interpreting the treaty or assessing a Party’s observance of it, with no power to award compensation.
Set against this is the 2024 India-UAE BIT, in force since August that year, which retains the investor’s right to sue and adds concessions no other recent treaty offers: portfolio investments protected, a category Brazil and India’s own Model BIT both exclude, and the local-remedies wait before arbitration cut from five years to three. The Ministry of Finance has confirmed a revised Model BIT is coming and will be ‘more investor-friendly’, without releasing a draft, so the UAE treaty is the only public signal of its likely direction. Two recent treaties negotiated by the same government therefore adopt substantially different approaches to investor protection.
The standard answer is that this is not inconsistency at all. States calibrate investment protection to leverage, and there is nothing unusual about an EFTA trade deal landing differently from a Gulf capital relationship that needed a replacement for an expiring BIT. While this explanation readily accounts for variations in the extent of protection, it leaves open the question whether it fully explains more fundamental differences in remedial architecture. Bargaining by leverage varies the price of protection: a longer wait before arbitration, a narrower definition of investment, a higher bar for compensation. It does not ordinarily remove the remedy altogether for four separate counterparties while installing a richer version of it, unseen elsewhere in Indian practice, for a fifth. If bargaining power alone explains the divergence, one might expect greater variation among the EFTA, Oman, Brazil and UK arrangements themselves. Instead, all four land on no investor remedy at all, with nothing on the record marking why the line falls there.
The WGIII episode matters because it surfaces the same unresolved choice where bilateral bargaining cannot explain it. A bilateral treaty can always be read as the product of that relationship’s particular leverage, so an EFTA negotiator and a UAE negotiator can each walk away with a coherent story even if the two stories contradict each other. WGIII removes that excuse, because non-disputing party submissions would apply however any individual treaty allocates rights between investors and states; India’s delegation was not protecting a bilateral bargain by opposing it. The interesting question is not that India drew on WTO experience, which states frequently do, but how that concern fits alongside India’s broader reform agenda. India opposed a transparency measure within a system that it was simultaneously arguing should undergo deeper structural reform. The multilateral setting also makes relationship-specific explanations less immediately available than they would be in a bilateral treaty negotiation.
None of this proves India’s treaty practice is incoherent in the sense of being arbitrary. Indian officials may well have a clear internal view of when investors should get a remedy and when they should not, and have simply never said what it is. Because no general framework has been publicly articulated, external actors may find it difficult to distinguish between deliberate differentiation and ad hoc variation. The forthcoming Model BIT revision is the first chance to remove it. If it adopts the UAE architecture, the EFTA, Oman, Brazil and UK treaties become the exceptions that need explaining. The forthcoming Model BIT revision presents an opportunity to clarify how the UAE treaty and the more restrictive approaches adopted elsewhere relate to one another. Either model could serve as the organising principle of India’s future treaty practice. What remains uncertain at present is whether the recent diversity reflects deliberate differentiation among partners or an evolving conception of investor remedies that has yet to be publicly articulated.
Apeksha Kachhawaha is an Associate, ASV Legal LLP, Delhi, India. Graduate (2026) from Maharashtra National Law University, Nagpur, India.
Kshitij Saruparia, Advocate, Rajasthan High Court, India. Graduate (2026) from NALSAR University of Law, Hyderabad, India.
Share:
With the support of
