Introduction
The investor-state dispute settlement system (‘ISDS’) has been an asymmetrical regime, often making investment arbitration an instrument of advancing the interests of investors of economically developed states. It has been argued that counterclaims by host States can serve as a potential balancing mechanism, making the regime less biased and also increase State’s confidence in investment treaty arbitrations. Article 46 of the International Centre for Settlement of Investment Disputes (‘ICSID’) Convention provides that tribunals can determine “counterclaims arising directly out of the subject-matter of the dispute provided that they are within the scope of the consent of the parties and are otherwise within the jurisdiction of the Centre.” While many countries, including India, are not a party to the ICISID Convention, Article 46 has nevertheless shaped the doctrinal architecture of counterclaims in investment arbitration. So, procedurally, counterclaims have been available within international investment arbitration for decades. However, states have rarely invoked it despite its longstanding availability. Even if invoked, rarely have tribunals established jurisdictions over counterclaims, and even if counterclaims overcome jurisdictional hurdles, they frequently fail on merits.
In a first in Indian BITs, the India-Uzbekistan BIT, entered into force in May 2025, has an express provision permitting Counterclaims by host States, as well as references to investor obligations. The provision appears to be a response to long-standing objections concerning consent and jurisdiction around counterclaims by explicitly indicating consent in the BIT. However, while the express counterclaim provision may be doctrinally innovative, I argue that it ultimately is demonstrative of the structural limits of counterclaims and entrenches the asymmetry of the ISDS. The following limitations continue to shape counterclaims: persistence of consent requirements despite explicit authorisation, close-connection requirements, and the normative disconnect between investor obligations and dispute settlement jurisdiction.
Consent and Treaty Interpretation
To establish jurisdiction over counterclaims in investment arbitration, ‘consent’ is a fundamental prerequisite. Under the traditional ‘arbitration without privity’ model, this consent manifests through the State’s offer to arbitrate in the BIT and the investor’s acceptance of the same through filing a request for arbitration. The question that arises then is whether this consent to investors’ treaty claims also extends to the State’s claims against the investor. Arbitral tribunals have read the consent in restrictive as well as expansive manners, requiring consent to be “found in the investment treaty itself” or “implied from the applicable rules” respectively. In Roussalis, exemplifying the restrictive approach, the tribunal opined that the basis of the consent must be found primarily within the text of the relevant BIT. Thus, it excluded jurisdiction over counterclaims filed by the State as the Romania-Greece BIT unambiguously contemplated only investor claims against states.
In his dissent in Roussalis itself, Professor Reisman challenged this restrictive reading by arguing that when states consent to ICSID jurisdiction, the consent component of Article 46 gets ipso facto incorporated in the resulting arbitration. This expansive view has found very limited success. Most tribunals continue to view the BITs as constituting the ‘outer limits’ of jurisdiction, implying textual supremacy of BIT’s for determining consent. In Gavazzi, adopting the Roussalis approach, the tribunal explicitly rejected the idea of inferring consent from the ‘spirit’ of the treaty and advocated for adhering strictly to the text. Thus, the historical failure of counterclaims is not because of its conceptual impermissibility, but because of tribunals insisting on consent being explicit in the treaty text and symmetrical regarding the obligations of the parties.
The Connection Requirement
After consent, the requirement that counterclaims bear a sufficiently close connection to the investor’s primary claim poses the second major hurdle. Article 46 mandates that counterclaims must arise directly out of the dispute’s subject-matter. However, there is a lack of a consistent standard for assessing this connection. Tribunals and scholars have divided opinions on whether a strict legal connection is required or merely a factual connection is sufficient. In Saluka v. Czech Republic, the tribunal insisted on the strict approach that the primary claim and the counterclaim should arise from the same legal instrument. The implication of this test is that counterclaims based on domestic law or contractual obligations would not satisfy the standard and, thus, would not be admissible. The standard has been critiqued for being ‘artificial and inefficient’ as it fails to reflect the reality of investment disputes which involve multiple legal orders. For instance, if an investor-claim allege expropriation of a mining concession under the BIT and a counterclaim is raised by the host State for environmental damage caused by the same mining operation under domestic environmental law or the concession contract, both claims arise from the same investment. However, a strict legal-instrument test would render the counterclaim inadmissible merely because it is grounded in a different legal source. Thus, in this process, the regulatory reality of investment projects, which routinely operate across domestic, contractual and international legal frameworks, is disregarded.
Certain decisions have also shown the possibility of a rather flexible factual connection standard, that is, the connection requirement is satisfied as long as both claims arise from the same investment operation. For instance, in Goetz v. Burundi, the tribunal held that there was a sufficient connection because both primary and counterclaims concerned the “same bank operating certificate”, which was the investment in dispute. A similar approach was taken in Urbaser wherein the factual nexus between the counterclaim and the investment was found to be sufficient to satisfy the connection requirement. This divergence in approach means that even explicit authorisation may prove ineffective if tribunals continue to adopt restrictive approaches to connection.
Notably, close connection is generally classified as an issue of admissibility rather than jurisdiction. Coupled with the uncertainty of the assessment standard, this allows for a possibility where the tribunals can exclude counterclaims that threaten to expand the scope of the dispute beyond the investor’s framing, even while formally acknowledging their permissibility.
Investor Obligations
A further requirement to succeed on a counterclaim is that States must demonstrate that investors have breached obligations owed to them. The issue here is that, on account of being asymmetric instruments, investment treaties, traditionally, impose obligations on States while only conferring rights on investors without any corresponding duties. So, without treaty-based obligations, States would need to look at alternative sources of law if they are seeking to bring successful counterclaims. Potential sources such as domestic law, international law and contractual obligations each have their own unique difficulties.
Firstly, while there is a general obligation on investors to comply with domestic laws, it does not automatically translate into a treaty obligation. For instance, in Amco v. Indonesia, the tribunal distinguished between obligations applicable to all persons under general law, tax fraud in this case, and obligations arising out of an investment agreement, ruling that, unless specifically elevated by treaty, general legal obligations fall to domestic courts.
Secondly, when it comes to international law, investors may be subject to certain negative obligations, but positive obligations requiring specific conduct remain largely absent. Acknowledging this limitation, the tribunal in Urbaser, while dealing with Argentina’s human rights-based counterclaims, noted that while states bear positive obligations to ensure access to water, there is no such corresponding positive duty on investors, and they only have a negative obligation to abstain from infringing the right.
Burlington and Perenco, the two cases where counterclaims have succeeded have both involved breaches of environmental obligations contained in petroleum extraction contracts. So, contractual obligations seem to offer a promising foundation, particularly in concession-based disputes. Yet, even in such cases, purely contractual obligations may not fall under tribunal jurisdiction, unless the treaty has a broad dispute resolution clause.
To rebalance the asymmetry of ISDS, a number of contemporary model BITs and International Investment Agreements, such as the SADC Model BIT 2012, the Morocco–Nigeria BIT 2016, and the Pan-African Investment Code 2016, have introduced express investor obligations. However, despite incorporation of such obligations, effective procedural mechanism for their enforcement continues to be a contested issue. It is against this evolving backdrop of “new generation” treaty design, aimed towards increasing the state’s right to control and hold investors accountable, that the India-Uzbekistan BIT’s express counterclaim provision must be situated.
The India-Uzbekistan BIT: Is it Solving the Right Problems?
In 2011, India received its first adverse award in White Industries under the Australia–India BIT, bringing attention to the BIT regime. Post this award, there was a surge of BIT cases against India, causing national concern and calls for curtailment of ISDS. India was forced to revisit its BIT regime, leading to the adoption of a Model BIT and termination of the old BITs. India’s new BITs are reflective of this recalibrated investment treaty policy. In India-Belarus BIT, one of the first BITs negotiated post adoption of the new Model BIT, the change began with the inclusion of a chapter on Investor Obligations for the first time.
It is in this context that India-Uzbekistan BIT was negotiated and signed. Chapter III of the India-Uzbekistan BIT also imposes investor obligations, including compliance with domestic laws and adherence to corporate social responsibility principles of labour, environment, human rights, etc. However, in contrast to the India-Belarus BIT which did not provide for a counterclaim provision, Article 16 in Chapter IV of the India-Uzbekistan BIT provides that “The defending party may initiate a counterclaim against the Investor or Investment for a breach of the obligations set out under Chapter II (or for breach of the object and purpose) of this Treaty…,” marking it as the next step in the evolution of India’s new-generation BITs. By expressly permitting counterclaims, it appears to address the formal consent objection that has historically plagued State counterclaims.
However, the provision in Article 16 limits counterclaims to breaches of ‘Chapter II obligations’ or for breach of the treaty’s object and purpose. The obligations laid down in Chapter II are substantive state obligations regarding treatment of investments, expropriation and compensation, transfers, etc. Chapter III, that lays down investor obligations, is not included in the counterclaims provision’s scope. So, investor obligations have not been framed as jurisdiction-triggering norms. This leads to a normative procedural disconnect: at the level of treaty text, obligations for investors exist, but they lack an independent procedural pathway for enforcement. The mere inclusion of investor obligations, thus, does not eliminate the inherent asymmetry of the non-reciprocal nature of investment treaty arbitration. The provision then only reinforces the derivative nature of counterclaims.
Although this counterclaim provision is a first in any of India’s BITs and does not feature even in the Model BIT 2016, the Draft text for the Model Indian BIT 2015 did contain a counterclaim provision. Article 14.11 of the draft text allowed states to initiate counterclaims against the Investor or Investment for breach of investor obligations that are laid down in Chapter III. Thus, it explicitly linked counterclaims to investor obligations, and it is precisely this connection that is missing from the India-Uzbekistan BIT’s Article 16.
The most immediate effect of Article 16 of the India-Uzbekistan BIT is that it resolves the consent ambiguity that plagued tribunals in cases like Roussalis and Gavazzi. By explicitly authorizing counterclaims in the treaty text, the treaty eliminates the need for tribunals to infer consent from procedural rules or the ‘spirit’ of the treaty. However, there are other consent related questions that persist. The fact that Article 16 is limited to the violation of obligations under Chapter II creates interpretive uncertainty. Do counterclaims alleging investor violations of environmental law have to be framed as a breach of the treaty’s object and purpose to invoke Article 16? Or would such a counterclaim be simply beyond the scope of the BIT?
Conclusion
The result of explicit inclusion of counterclaim provision in the India-Uzbekistan BIT is that the decisive filtering function is relocated from jurisdictional text to the discretionary realm of admissibility; from rule-based exclusion to standard-based control, specifically the ‘direct connection’ requirement. Instead of limiting it, this shift entrenches the discretionary space for arbitrators. The classification of the requirement of close connection as a question of admissibility, rather than jurisdiction, allows tribunals significant flexibility. While the tribunal may acknowledge consent in principle based on the existence of Article 16 of the BIT, they can still exclude specific claims by citing insufficient alignment with the investor’s claims. Such exclusion would not be exceptional but systematic, reflective of the reluctance to transform ISDS into a forum for regulatory enforcement. Thus, while counterclaim offers the possibility of rebalancing the system, the mere inclusion of express counterclaim provision, such as in the India-Uzbekistan BIT, does little to disrupt these systemic incentives and the structural imbalance that defines the regime.
Ritu Ranjan is a fourth-year law student at the National Law School of India University, Bengaluru.