Announcements

IIHS University

The IIHS (Institution Deemed to be) University is now accepting applications from interested candidates for the second batch of their Master’s programmes. The University offers a set of five transformative interdisciplinary programmes for the 2026-27 intake.

• Master of Science in Sustainability Science and Practice.
• Master of Science in Climate Change Science and Practice.
• Master of Science in Urban Economic and Infrastructure Development.
• Master of Arts in Urban Studies and Practice.
• Master of Arts in Human Development Policy and Practice.

The attached brochure gives an overview of the five programmes. Interested candidates can apply at www.iihs.ac.in; applications are open until 20 March 2026.

The University also offers a PhD in Urban Studies and Practice, applications for which will open on 13 March 2026.

As India undergoes the largest rural-to-urban transition in human history, urbanisation is fundamentally reshaping its economy, society, culture, and environment. Recognising this defining moment, the University is premised on the need to transform the current nature of urban education, one of the most important drivers for India’s national development and sustainable global futures.

Recent graduates and young professionals from any educational background are eligible to apply to any of the five Master’s programmes. For example, even bachelor’s degree holders in non-science subjects can apply to IIHS University’s Master of Science programmes.

For queries, you can write to admissions@iihs.ac.in or contact +91 99012 55788, 96325 20741 (10 am to 6 pm India Time, Monday to Friday). IIHS, Sadashivanagar Campus, Bengaluru 560 080, India and IIHS, Kengeri Campus, Bengaluru 560 060, India.

Author: Siddarth Raman

Siddarth Raman is a senior research lead at XKDR Forum.

On this blog:

• High-Voltage Treatment for a Comatose Elephant, 28 December 2025.
• Establishing the baseline for the Kollam district court reform, 21 November 2025.
• Beyond Pendency: Counting Cases Correctly, 5 October 2025.
• Get them to the court on time: bumps in the road to justice, 12 June 2025.
• Who is litigating cheque bounce cases?, 5 June 2023.

Beyond Syndication: Unlocking the Power of Single-Asset Securitisation

by Amrita Agarwal and Harsh Vardhan.

An overlooked yet transformative pillar of India’s evolving securitisation framework is the newfound empowerment of lenders to undertake the securitisation of a single, standard asset (Master Direction – Reserve Bank of India (Securitisation of Standard Assets) Directions, 2021. RBI/DOR/2021-22/85, September 24, 2021). To appreciate the magnitude of this shift, one must first look at the historical contours of the Indian market. Traditionally, securitisation has been synonymous with "pass-through certificates" (PTCs) backed by granular pools of consumer debt—typically vehicle and personal loans or microfinance receivables. The market has also frequently, if somewhat inaccurately, applied the term to "direct assignments," where loan portfolios are sold outright between lenders.

Indian securitisation has long been a predictable, if somewhat narrow, affair. Historically, the market has been the domain of "pass-through certificates" (PTCs)—complex bundles of granular consumer debt, ranging from tractor loans in Punjab to microfinance receivables in Tamil Nadu. The logic was simple: there is safety in numbers. By pooling thousands of small loans, lenders could hedge against individual defaults, creating a diversified product for investors.

In September 2021, the Reserve Bank of India (RBI) issued a comprehensive new framework. The impact on traditional volumes was immediate. PTC transactions, which languished at a modest ₹0.33tn in the 2021 financial year, surged beyond ₹1tn by 2024. Data from the first half of 2025 suggests the market is reaching critical mass, with volumes already beyond ₹0.7tn.

Yet the true significance of the 2021 guidelines lies not in the scaling of the old model, but in the birth of a new one. In a radical departure from the "diversified pool" orthodoxy, the new regime permits a lender to carve out a single, standard project loan, house it in a bankruptcy-remote trust, and issue securities against that solitary asset.

It is the "securitisation of one" -- a sweet insight that is one pioneer away from becoming a market reality.

Breaking the CLO Mould

Globally, the securitisation of commercial credit is a mature, if occasionally notorious, discipline dominated by Collateralised Loan Obligations (CLOs). But the Indian iteration is a distinct breed. A typical European or American CLO might bundle 150 different corporate loans to achieve the "granularity" required for a triple-A rating.

India’s provision, by contrast, allows for the atomisation of a single, massive exposure. This offers banks unprecedented agility in managing the long-tenure, high-ticket exposures inherent in infrastructure finance.

Until now, such projects were managed through the consortium model: a cumbersome, 20th-century process where a lead bank manages a gaggle of lenders or "down-sells" portions of the debt to stay within internal risk limits. Crucially, under syndication, the loan remains an illiquid fixture on the balance sheet. Single-asset securitisation changes the chemistry of the transaction, turning a private contract into a tradable financial instrument.

Efficiency by design

The single-asset model introduces a surgical approach to capital efficiency. Under RBI rules, a fully disbursed project loan becomes eligible for securitisation after a six-month "minimum holding period." This timing is strategic.

In the volatile world of infrastructure, early-stage risks -- regulatory bottlenecks, environmental clearances, and land acquisition disputes -- are at their most acute during the first shovel-load of dirt. By the time a loan is fully disbursed and has survived its first six months, these foundational uncertainties have typically receded.

Securitising at this juncture allows a bank to capture the higher yields associated with the high-risk inception phase, only to exit and free up capital just as the asset settles into the boring, predictable cash flows of a stable utility.

Beyond individual relief, this mechanism addresses two systemic vulnerabilities that have long haunted the Indian subcontinent:

1. The Asset-Liability Mismatch: Indian banks are perennially caught in the "borrow short, lend long" trap: using three-year deposits to fund 20-year power plants. Securitisation provides a vital exit valve, moving long-dated assets off the books.
2. Broadening the Investor Base: By transmuting a private loan into a security, the industry can tap into the deep pockets of institutional liquidity. Insurance companies and pension funds, which crave long-duration cash flows, have historically lacked a direct route to project-level credit without taking on the broader, often messy, corporate risk of the developer.

The friction points

If the advantages are so compelling, why has the market not yet ignited? The answer lies in a combination of dormant private capital expenditure and significant fiscal friction.

For the better part of the last decade, India’s infrastructure story has been a public sector affair, financed largely through the bond markets. As the private sector begins to re-engage, two primary hurdles remain:

The stamp duty trap

The primary hurdle is the prohibitive cost of registration. In many Indian states, transferring a loan to a trust can incur charges of between 3% and 4%. While some states have capped this at 1% for pooled assets (creating regional hubs for securitisation) the single-asset model remains burdened by archaic fees. Analysts argue that state governments must act in their own enlightened self-interest to reduce these frictions if they wish to see infrastructure projects in their backyard funded efficiently.

The Regulatory Glass Ceiling

Current Securities and Exchange Board of India (SEBI) regulations present a barrier. Rules for listed securitised debt instruments require that no single obligor represents more than 25% of the asset pool. This effectively bans single-asset securities from being listed on recognized exchanges.
This lack of listing creates a domino effect. Insurance companies, governed by strict IRDAI mandates, are generally restricted to "approved investments" that must be listed. Without a SEBI carve-out, the senior, high-quality tranches of these deals (precisely what insurers want to buy) remain off-limits. Furthermore, under Minimum Retention Ratio (MRR) rules, the originating bank must keep a 10% "skin in the game," usually the junior "equity" tranche. This leaves the "Senior Tranche" looking for a home that current listing rules effectively block.

How this fits into Indian economic growth

As the private capex cycle begins its long-awaited ascent, the demand for sophisticated financing tools will be immense. The "securitisation of one" is no longer a mere regulatory curiosity or an academic exercise. It is a vital instrument for a banking sector that must fund a nation's growth without choking on the resulting long-term risk. For the pioneer who manages to navigate the stamp duty and the listing hurdles, the rewards and the market share will be substantial.

The framework is there. The assets are coming. All that remains is for the market to move beyond the safety of the pool and embrace the power of the one.

The authors are experts on finance and public policy.

Are all regulators equal? The Supreme Court doesn't think so

by Chitrakshi Jain and Bhavin Patel.

Introduction

The judiciary has played a very important role in the development of the regulatory state in India. Thiruvengadam and Joshi (2013) argue that this is dissimilar to the evolution of the regulatory state in the global North, where courts have been marginal actors. In contrast, the Indian Supreme Court, they reveal, has on multiple occasions mediated conflicts that arose during the reform of telecom regulation and in this process imparted institutional credibility to the sectoral regulator. Consequently, the Supreme Court's role in designing the regulatory state should be studied more closely. In this article, we examine the implications of select Supreme Court decisions which shape the characterisation of regulators' functions and their appellatory rights.

It bears repeating that regulators in India accumulate executive, legislative, and quasi-judicial functions. This accumulation of functions combined with ambiguous laws that outline the powers of regulators has led to a scenario where the courts have been asked to clarify the character and nature of functions and the corresponding appellate remedies.

Regulators in India are not similarly situated vis-a-vis the powers that they have been bestowed with. This complicates matters even further. In the case of utilities, regulators are tasked with the crucial function of determining tariffs (amongst others). The allocation of this function to regulators was critical in ensuring that these sectors are depoliticised, for state governments perceivably manipulate rational determination of tariff to advance their political interests. Regulators are instead expected to rely on their expert knowledge and make tariff determination a rational and efficient exercise which protects consumer interest.

In three decisions, PTC v CERC (2010), GRIDCO v Western Electricity Supply Company of Orissa Ltd and Ors. (2023), and AERA v Delhi International Airport Ltd. (2024), the Supreme Court has been asked to rule on the nature of tariff determination in regulators, especially the various Electricity Regulatory Commissions (ERCs). In this article, we examine how these decisions have shaped regulatory practices and study whether the interventions by the Supreme Court have:

1. been consistent in identifying the nature of tariff determination across regulators;
2. led to inconsistency in the substantive appellate rights of regulated entities in sectors governed by different regulators.

We first outline the reasoning from the recent decisions of the Supreme Court on tariff determination. We proceed to identify the challenges they create for the development of regulatory theory in India generally. Lastly, we conclude with the suggestion to legislate based on principles that should apply uniformly across regulators.

Judicial decisions

The PTC, GRIDCO, and AERA decisions have a bearing on the following issues:

1. What is the nature of tariff determination? Is it a quasi-judicial function?
2. Can a regulator prefer an appeal against an adverse appellate order which modified or overruled the regulator's tariff order?

In PTC, the Supreme Court was asked to clarify whether APTEL has the power of judicial review and the authority to decide on the validity of regulations made by the Central Electricity Regulatory Commission (CERC). In the course of its reasoning, the Court observed that tariff determination, while being legislative in character, is made quasi-judicial in the context of the CERC because the Electricity Act, 2003 (Electricity Act) has made tariff orders appealable to APTEL. The judgment was rendered by a constitution bench (5 judge) of the Court and limited the scope of its application to the statutory scheme in the Electricity Act. This has not stopped parties from invoking the reasoning given in PTC in matters where other regulators are involved. For example, counsels for Delhi International Airport Ltd., in AERA placed reliance on PTC to argue that tariff determination is a quasi-judicial function.

In GRIDCO, a division bench of the Supreme Court developed upon the reasoning in PTC (that tariff determination is a quasi-judicial function) and extended it to argue that it would be improper for regulators to prefer appeals against orders of the APTEL to the Supreme Court. While the Court did not sufficiently explain its reasoning, it implies that regulators cannot be parties to appeals against orders which are passed while discharging a quasi-judicial function.

Finally, in AERA, a full bench of the Court clarified the law on impleadment of regulators as parties in appeals. It held that while an authority discharging a quasi-judicial function should not be impleaded as a party to an appeal, such an authority must be impleaded as a respondent in an appeal against its order if it was issued in exercise of its regulatory role or if its participation can further effective adjudication. The Court said that the PTC judgment should be treated as authoritative only to argue that classifying tariff determination as legislative or an adjudicatory function should depend on the statute which distributes the powers to the relevant authority. To hold that AERA is allowed to appeal, the Court after reading the statute argues that tariff determination for aeronautical services is a 'regulatory' function.

Collectively read, these judgments fall short of clearly articulating an objective criteria for demarcating the different functions exercised by regulators. Consequently, these determinations have to be located in the laws that create each individual regulator, and post facto by judicial interpretation. The lack of a well-defined objective criteria for identification of different functions has serious implications for Indian regulatory theory.

Problems for regulatory theory

These cases raise three major problems for Indian regulatory theory:

SRAs are treated differently

While PTC and GRIDCO hold that tariff determination by ERCs is a quasi-judicial function, AERA holds that a similar exercise by the regulator would be a 'regulatory' function. All three cases rely on interpretations of the SRAs' parent statutes to arrive at this conclusion. The judicial discussion hinges on points such as whether the statute provides an appellate mechanism against tariff determination orders. But this does not answer the question of what the core characteristics of quasi-judicial or regulatory functions are. This is ad hoc and intricate statutory interpretation which does not illuminate any underlying principles of Indian regulatory jurisprudence.

While GRIDCO casts doubts on an ERC's ability to file an appeal against an adverse APTEL order on an appeal from its quasi-judicial orders (which, as we have noted above, include tariff determination orders), AERA can file appeals against TDSAT orders arising out of its tariff determination orders. This creates an arbitrary distinction between these two SRAs. The court decisions turn on intricate statutory interpretation, but do not tell us anything about the distinguishing features of the underlying nature of tariff determination at these two SRAs.

No ex ante certainty on nature and appealability of orders

Since no clear rules emerge from these decisions regarding how to distinguish between quasi-judicial and regulatory orders, or whether SRAs can file appeals against adverse appellate tribunal decisions in appeals against their quasi-judicial orders, there is no clarity on these points as regards other SRAs. Whether a particular order passed by an SRA is quasi-judicial or regulatory, or whether it can file an appeal to defend such orders, has to be decided on a case-by-case basis. The answers to these questions can only obtain certainty when the Court pronounces on them.

This means that there is no ex ante conclusive answer to these questions in Indian regulatory jurisprudence. SRAs, regulated entities, and other concerned parties have a tough enough job navigating the maze of Indian regulatory case law as matters stand - they do not need the added uncertainty created by these decisions to make things worse. It is also deeply concerning that fundamental questions about character of functions and right to appeal are being decided decades after these laws were enacted.

The lack of ex ante clarity on these matters also complicates the job of the appellate tribunals, who have to justify choosing between precedents that are not consistent with each other. For example, very recently APTEL had to carefully navigate its way through PTC, GRIDCO and AERA to justify placing reliance on AERA and argue that tariff orders are also legislative and regulatory in character and that regulators should be allowed to defend their orders.

APTEL becomes the final authority on tariff determination

Tariff determination is delegated to ERCs via legislative mandate. The reasons traditionally offered for this include sectoral expertise and independence from the political executive. In addition, we may note that ERCs are much closer to the ground, as it were, and have (or should have) access to much more information about factors for tariff determination. As such it is ERCs, and nobody else, who should determine tariffs.

But if an ERC passes a tariff determination order, and APTEL sets it aside or modifies it, then the ERC cannot file a second appeal before the Supreme Court, even when they are the only contesting parties. This effectively means that APTEL's decision on tariff determination automatically trumps the decision of the expert body specifically identified by Parliament to take such decisions. Additionally, if APTEL makes an erroneous decision, the possibility of it being corrected is diluted unless regulated entities challenge it before the Supreme Court.

This amounts to an egregious violation of the principle of separation of powers across Parliament and the judiciary.

Conclusion

The problems identified above all point to the need for a better theory of institutional design for Indian regulators. One can take the view that this is still a relatively new area of Indian jurisprudence, since regulators are (other than the RBI) a fairly modern addition to the structure of the Indian government. That said, the ERCs were created in 1998, and it has been 28 years since. Similarly, it has now been well over three decades that SEBI was established, and over a quarter of a century since TRAI was set up. Surely enough jurisprudential mileage has been accumulated now to warrant some stock-taking:

1. We need a clear articulation of the different types of functions that regulators may discharge, and a categorisation of these functions into the three broad heads of executive, quasi-legislative, and quasi-judicial. Such a theoretical categorisation is necessary because judicial interpretation has led to varying, ad hoc categorisation across SRAs. Without a strong justification - which cannot be based on statutory hair-splitting - this is arbitrary, since it means that regulated entities in different sectors are treated differently. Calling a function 'quasi-judicial' means something: appellate rights exist, and certain judicial procedures have to be followed in discharging them; calling it 'executive' means limited grounds of judicial review are available through writ proceedings, and fewer procedural safeguards apply.
2. The Tinbergen Rule may help here - for each policy target a regulator is expected to influence (tariffs, enforcing regulation, contractual disputes, etc.), they must use a distinct instrument. This way each policy target stands a better chance of being achieved, and no conflicts arise from attempts to use the same instrument to achieve multiple targets. Law makers will have to think harder about the specific targets they want regulators to achieve, and for each, describe the instrument they must use. This legislative prophylaxis eliminates the risk of contradictory judicial interpretations later.
3. Pertinently, in its 31st report, the Parliamentary Standing Committee on Energy reviewed the Electricity Bill 2002 and acknowledged that assigning all the three functions to the same institution can create challenges and had asked for a reconsideration of the clauses assigning functions to the regulators. It is unclear whether a careful examination of the relevant provisions has since been undertaken; however, the implications of not delineating the functions in the statute itself can be seen in the repeated litigations before the courts.
4. Tariff rationalisation is listed as an objective in the long title of the Act. Regulators have expertise and independence from the executive, hence they are entrusted with such functions in infrastructure industries like electricity. Regulators are also obliged to ensure costs are efficiently allocated and consumer interest is protected at the same time. If we expect regulators to be accountable, they should also have the right to defend their decisions on tariff determination in a second appeal; not allowing this prevents them from discharging their function effectively.
5. Given the need for a uniform theory across regulators, and the need to avoid ex post determination through the judicial machinery, the categorisation of functions we propose should be done through a statutory instrument. The American Administrative Procedure Act is one example of how this can be done. It is possible that new functions emerge with changing circumstances; these can be addressed through amendment, rather than judicial review.

These measures will help achieve the goals of uniformity across regulators, non-arbitrariness in classification of functions, and greater certainty and predictability for regulated entities. The current proliferation of judicially-determined, ad hoc rules is undesirable, and helps with none of these objectives.

References

Judiciaries as Crucial Actors in Regulatory Systems of the Global South: The Indian Judiciary and Telecom Regulation (1991-2012) by Arun K Thiruvengadam and Piyush Joshi [2013] Oxford University Press.

Electricity Tariffs by Javier Reneses, María Pía Rodríguez and Ignacio J.P. Arriaga [2013] Springer.

Chitrakshi Jain and Bhavin Patel are researchers at the TrustBridge Rule of Law Foundation and would like to thank Renuka Sane and Ajay Shah for useful feedback.

Cholesterol in the Indian GST

by Arbind Modi and Ajay Shah

In 2017, the Goods and Services Tax (GST) replaced the fragmented indirect tax structure of excise duties, service tax and state-level value added tax (VAT) with a unified, destination-based value-added tax. GST was conceived in Kelkar 2003, as an Indian adaptation of the global idea of the value added tax, which has the desirable features of having no cascading effects and being neutral to international trade.

Input tax credit is the beating heart of the GST: the tax paid on inputs must be credited against output tax liability. In a new XKDR Forum Working Paper, Input Tax Credit and refunds under GST in India: Conceptual and legal framework, we show that many features of the modern Indian GST interfere with input tax credit. In this, we combine the strategic sense of modern public economics coupled with a detailed reading of the law.

We worry that the Indian GST is a case of isomorphic mimicry: it resembles the external form of a modern VAT in form but lacks its underlying function. While Section 16 of the CGST Act establishes foundation of ITC entitlement, Section 17 narrows the scope through apportionment and blocked credits. Section 17(5) explicitly denies ITC for business inputs such as motor vehicles and works contracts. Furthermore, the refund framework under Rule 89 limits refunds to Net ITC and embedding revenue safeguards at the expense of neutrality. Critically, the law excludes capital goods and services from refund calculations in these cases. This breaks the credit chain.

When credit is denied, the tax on inputs becomes a cost. This creates a cascading tax through the supply chain. The tax system functions as a tax on intermediate production rather than on consumption. By blocking input tax credit for capital goods, the GST acts as an incentive for reduced investment.

This is not how the value added tax works internationally. India stands as an outlier relying on a narrow refund design, heavy compliance preconditions, and extensive exclusion. The current system minimizes administrative risk by denying refunds ex-ante, rather than managing risk through ex-post verification. The recent movements on `GST 2.0' do not deliver an improved GST as they actually worsen the blockage of input tax credit.

These flaws in tax policy have harmful effects for India: (a) Reduced capital deepening and reduced investment that hampers the emergence of high productivity and high wage firms; (b) Unfair competition from imports against domestic producers; (c) Reduced export competitiveness. The paper proposes a reform path focused on three pillars: guaranteeing full flow-based ITC across all inputs, rationalising rates to a single band, and automating refund administration.

The authors are researchers at XKDR Forum.

High-Voltage Treatment for a Comatose Elephant

Unshackling the Elephant by Anand Prasad: A Review

by Siddarth Raman.

Anand Prasad's Unshackling the Elephant provides an interesting clean-sheet critique of the Indian legal system. India's development story into the 21st century is shackled by a judicial system that is stuck in the past. The book succeeds incredibly well at applying economic logic and management principles to the judiciary. Many of the suggested reforms have the weight of common sense - the kind that becomes visible once someone has taken the trouble to articulate it well. It stumbles into uncertain territory when suggesting we succumb to indigenous instinct, especially given the fragility of India's current state capacity and my own preference for cautious reforms that avoid the treacherous currents of populism.

This book matters. It is precise in its diagnosis and bolder in its questions than most will venture. Systems rest on assumptions we stop interrogating; surfacing these debates is important because knowing why we got here tells us what needs to be fixed and what doesn't need relitigating.

The Low-Hanging Fruit of Process Modernisation

On several fronts, Prasad's prescriptions will invite broad agreement - particularly from those who have witnessed the impact of process modernisation in other fields and wondered why the legal system has resisted it for so long.

The law is both judicial reasoning and process management. Other fields have walked the journey of process improvement. IT services, consulting, accounting, even corporate law in some aspects. Templatised pleadings and standardised procedures need not be recreated from first principles for every matter; translation tools chip away at language barriers; and the knowledge management systems Prasad envisions - his "legal knowledge grid" - would make legal corpora accessible to law graduates and large language models alike. Ideas of virtual hearings and asynchronous proceedings push this further, reducing the friction of physical presence without necessarily compromising procedural integrity.

These are safe bets. They enhance capacity without demanding discretionary overreach. Early efforts have begun and should be embraced (see here, here, here); many of these ideas deserve serious piloting as part of the eCourts Phase III modernisation effort.

Prasad also questions the figure of the "all-purpose judge" -the expectation that foundational legal education alone equips one to adjudicate matters spanning technology, finance, and specialised commercial arrangements. The assumption may have held in a simpler era. Whether it holds today is less certain.

A note of caution on artificial intelligence. The justice system functions because society accepts the state's monopoly on coercive power and consents to have it applied through due process. This is a social contract grounded in human accountability. AI may assist judges. But the decision must remain with someone who can be questioned, overruled, and held responsible. As much as the engineer in me might wish law were code, it is not.

Fixing the Incentive Systems

Courts are not just forums for dispute resolution. They are markets where incentives shape behaviour.

Consider the interest rate regime applied to legal disputes. Courts award simple interest at rates that don't match any reasonable cost of capital. The effect is predictable: delay becomes a strategy. A defendant who owes money has every reason to stretch proceedings; time works in their favour. Prasad is correct to argue that judges need to understand the time value of money. Aligning judicial interest rates with economic reality would shift the calculus against strategic delay.

On damages, Prasad makes the case for stronger deterrence -particularly through punitive awards for corporate fraud. On costs, he argues that allowing successful litigants to recover expenses would reduce barriers to access, enabling those confident in their position could invest in quality representation without bearing the full risk.

He also correctly identifies information asymmetry and power imbalance as distortions in the current system. Those with deeper pockets hire better lawyers, and in an adversarial system, this matters considerably. His proposals - removing restrictions on contingency payments, allowing lawyers to advertise, and litigation funding - would make legal representation function more like a market, creating a more level playing field.

Like Chief Justice Suryakant, Prasad echoes the need for judicial performance reviews. This deserves serious consideration, though any implementation must balance accountability against judicial independence. For the bar, he calls for high penalties for malpractice and stricter consequences for perjury. Whether the profession will police itself remains an open question. Prasad's experience understanding lawyers and litigants shines through in this section - he dissects how courts, like any market, respond to incentives.

The Temptations of an Overhaul

Anand Prasad's frustration with the legal system is understandable. Stepping into a courtroom feels like walking back through time.

While some structural reforms like splitting the Supreme Court into a constitutional court and court of appeals or curbing judicial legislation merit serious debate, some of the more radical proposals warrant caution. Moving towards an inquisitorial system would transform judges from neutral adjudicators into active investigators. In a country where discretionary power is prone to abuse, such a shift is a dangerous gamble.

The most troubling temptation is the urge to decolonise the system. Weakening foundational principles like the presumption of innocence in favour of indigenous traditions risks legitimising majoritarian sentiment as law, especially when honour killings persist in the 21st century and extrajudicial encounters meet public approval. The philosophical questions in attempting an Indic reinterpretation are formidable - attempting to balance ideas of karma against retributive justice, uniform civil codes against community specific practices, or the contextual obligations of Raja Dharma against the common law tradition that all stand equal before the law.

The Contradictions

There is an internal tension in the book's vision. It praises codification for its clarity and bemoans lack of consistency in judgements while simultaneously advocating for inquisitorial discretion and culturally-responsive justice. Codification demands predictability. Instinctive justice invites its opposite. A similar tension runs through the hope that algorithms will fix what humans could not. AI may detect patterns across thousands of judgements, but it cannot bear responsibility for any one of them.

The deepest contradiction is one of trust. The book hopes to build a high-trust society where precedent holds and contracts are sacred. This is hard to reconcile with privileging cultural instincts that are fluid and contested. One cannot ask people to trust in precedent while empowering judges to override it.

Conclusion

The boldness of the project deserves recognition. Unshackling the Elephant is a precise diagnosis of a system that has resisted reform for too long. Many treatments align with modernity. But there is a rebel streak - perhaps born of frustration - that carries risk. Giant shocks may end up killing the elephant rather than reviving it.

References

Prasad, Anand. 2025. Unshackling the Elephant: Transforming Indian Law, Culture and Economy. Bloomsbury India.

Siddarth Raman is Senior Research Lead at XKDR Forum.

Can technology augment order writing capacity at regulators?

by Natasha Aggarwal, Satyavrat Bondre, Amrutha Desikan, Bhavin Patel and Dipyaman Sanyal.

Indian regulators have extensive quasi-judicial powers that they express through adjudicatory orders. It is critical that these powers are exercised in a proportionate, legitimate, and well-reasoned manner, as they not only impact the persons directly involved, but also the wider ecosystem in which they operate. Arbitrary actions, unsubstantiated by clearly articulated reasoning, can raise serious concerns around the legitimacy of regulatory actions and lead to a loss of confidence in the regulator. Such actions may also be set aside by appellate and review fora. Clearly written, well-researched, and reasoned orders help provide clarity, predictability, and knowability of the law, which are key indicators of a rule of law system (Aggarwal, Patel and Singh, 2025). Our study of the state of Indian regulatory order writing shows there is room for improvement in this regard.

We notice a growing interest in the use of Generative Artificial Intelligence (Gen AI) to resolve procedural inefficiencies at quasi-judicial and judicial authorities in India (Supreme Court Committee on AI, 2025; Kerala High Court, 2025), coupled with concerns around the potential dangers of using such technologies without adequate safeguards. Against this background, in a new working paper titled, 'Can technology augment order writing capacity at regulators?' we critically examine the opportunities and challenges of using technology, in particular Large Language Models (LLMs), to assist regulatory order writing in quasi-judicial settings.

The paper proposes augmenting rather than replacing human decision-makers, aiming to improve regulatory order writing practice through responsible use of LLMs. It identifies the core principles of administrative law that must be upheld in these settings - such as application of mind, reasoned orders, non-arbitrariness, rules against bias, and transparency - and analyses how inherent limitations of LLMs, including their probabilistic reasoning, opacity, potential for bias, confabulation, and lack of metacognition, may undermine these principles.

While the available Indian literature on the topic focuses largely on these limitations, and on critiquing proposals based on an over-reliance on technocratic means to improve state capacity, this paper's contribution lies in its integrative work: we draw upon the design principles articulated in frameworks developed in other jurisdictions and relate them to the applicable principles of Indian administrative law. We use this synthesis to develop a Problem-Solution-Evaluation (PSE) framework that is attentive to international practice, the legal principles underpinning quasi-judicial decision-making in India, and problems and limitations inherent to GenAI and LLMs.

The PSE framework proposed in the paper maps specific technical, design, and systemic solutions to each identified risk, and outlines evaluation strategies - end-to-end, component-wise, human-in-the-loop, and automated - to ensure ongoing alignment with legal standards. An overview of the framework is set out in the table below:

Table 1: Applying the Problem-Solution-Evaluation framework. This table illustrates how the PSE framework can be operationalised to align the design, development and use of LLMs for order writing assistance with the requirements of Applicable Law.

Problem	Applicable law	Solution	Evaluation
Non-application of mind	Non-application of mind; Failure to provide reasons; Arbitrariness	Interface Checkpoints; Confidence Score Display; Dual-Prompt Pipelines; Functionality Limitation; Constraint Enforcement; Workflow Design for; Review Role-Based Access	Edit Rate; Turnaround Time (TAT); Prompt Divergence Rate; Coherence Score
Black-box problem	Failure to provide reasons; Transparency	Chain-of-thought prompting; Input Token Influence Identification Symbolic Reasoning Systems Traceability tools; Visualisation; Simplified model explanations	Clarity rating; Audit Trail Incidence Document Traceability Rate
Potential for bias	Rules against bias; Arbitrariness	Data Preprocessing; Bias penalisation; Domain-specific content filters; Automated Bias Flagging Tools; Establishment of Legal Fairness Criteria; Mandatory Periodic Benchmarking	Bias Flag Rate Override Percentage; Fairness Benchmark Scores
Confabulation problem	Non-application of mind; Failure to provide reasons; Arbitrariness	Retrieval Augmented Generation; Post-Generation Verification; Legal Knowledge Graph Integration; Mandatory reviewer verification; Watermarking for traceability; Communicate technical limitations	Secondary LLM ''Judge'' for Fact-Checking; End-to-End Evaluation Tools Hallucination Rate; Retrieval Precision@k/ MRR NLI Coherence Checks; Self-Consistency Rate
Lack of metacognition	Non-application of mind; Arbitrariness	Prompt engineering; LLM as a judge; Iterative improvement from feedback	Closeness Metric; Human evaluation on overconfidence in output
Training corpus	NA	Adaptive Scraping Frameworks; Sector-specific pre-training; Structured Entity; Extraction and Legal Knowledge Graphs; Isolated Model Containers; Source inclusion; Perplexity tracking; Legal Retrieval Benchmarking; Curate sector-specific legal databases	Crawl coverage; OCR Error Reduction; Validation perplexity; Retrieval lift
Data security and privacy	NA	Stringent access control; Synthetic supervision-based PII detectors; NLP filters for information masking; Isolated Model Containers; On-premise infrastructure	Unauthorised access attempts; Mean Time To Remediation (MTTR); Penetration Test Pass Rate; PII Detection Accuracy

By itself the framework may be insufficient. It must be supplemented with systemic measures taken at the regulatory level. We offer stage-wise recommendations on how LLM-based order review tools can be built for and used in regulatory adjudication.

References

Natasha Aggarwal, Bhavin Patel, and Karan Singh, "A Guide to Writing Good Regulatory Orders" [2025] Trustbridge Rule of Law Foundation Working Papers.

Anurag Bhaskar and others, "White Paper on Artificial Intelligence and Judiciary" Centre for Research and Planning, Supreme Court of India, 2025.

High Court of Kerala, "Policy Regarding the Use of Artificial Intelligence (AI) Tools in District Judiciary" Official Memorandum HCKL/7490/2025-DI-3-HC Kerala, 2025.

Natasha Aggarwal, Amrutha Desikan and Bhavin Patel are researchers at the TrustBridge Rule of Law Foundation. Satyavrat Bondre and Dipyaman Sanyal work on AI and technology at dōnō consulting.