This post is part of Notice & Comment’s symposium on Peter Conti-Brown and Sean Vanatta’s Private Finance, Public Power: A History of Bank Supervision in America. For other posts in the series, click here.

Private Finance and Public Power by Peter Conti-Brown and Sean Vanatta offers the comprehensive account of how financial regulation intersects at the frontlines with those that it seeks to regulate. This is an extraordinarily powerful project that speaks to the discipline of financial regulation and explains why it is so much more than an arcane body of rules. Rather it also represents the blood, sweat, and toil entailed in the act of supervision. Perhaps its most significant contribution lies in its forensic exposition of the mechanisms by which informed discretion governs supervision, how such judgment has infused oversight across modern U.S. banking history and the way in which it has shaped the substantive delivery of financial regulation.  Conti-Brown and Vanatta’s book already reveals the seeds for its second edition. In particular, this work prompts urgent questions surrounding how the exercise of public power vis-à-vis private capital will evolve as digitization in finance becomes ever more deeply embedded into the delivery structures for bank supervision. What will happen to discretion? How will this change the relationship between public and private in market regulation? Where banking supervision entails more substantive deference to “black box,” artificially intelligent processes, what becomes of the trust traditionally placed by the public and bank supervisors if this confidence might have helped underpin subsidies advanced to banks to engage in maturity transformation and financial engineering? Is banking supervision entering a brand-new era?

The acceleration of digitization across financial markets raises at least three major challenges for the kind of discretion-driven supervision that has formed the subject matter for Conti-Brown and Vanatta’s project. First, bank supervisors are confronting a financial landscape where essential services are increasingly disintermediated away from the banking sector without ever fully leaving bank balance sheets. Take the case of payments. Non-bank payment services, delivered especially appealingly in the wake of COVID-19’s transition to greater cashlessness, has fueled the uptake of technologies like non-bank issued digital wallets (e.g., Venmo or CashApp). In 2023, Venmo processed $276B in total payments volume. Stablecoins, or tokenized representations of a U.S. dollar, have exploded in issuance from essentially nothing in 2019 to over $200 billion outstanding in mid-2025. Despite the de-emphasizing of banks as key payment services providers, however, they remain central to the plumbing of payments by providing the deposit accounts where non-banks warehouse their cash as well as through which they access major USD payment schemes such as ACH and Fedwire. This extension in the supply chains for payments to place banks slightly further along the periphery without removing them altogether places supervisors in a challenging position. To understand the exposure that any bank might face, supervisors also have to consider the performance and risk profile of the various payment services providers that need to park their funds in a bank deposit account. Bank supervisors have to flex technological expertise to understand the dynamics of more novel stablecoin markets, for example, or to consider the ways in which applications such as digital wallets might be vulnerable to risks such as hacks and thefts. Such learning and technological evolution might underscore the continuing importance of supervisory discretion. Yet, deeper digitization, extensive data generated by payment applications, and the complexity of a more fragmented payment landscape, alongside varying levels of familiarity with new technologies, can also encourage greater reliance on more automated, artificial intelligence supervisory mechanisms that aid assessments in modern day financial systems.

Secondly, bank supervisors are facing a challenge from the rise of decentralized systems into which banks are dipping their toes to test out new products and markets. Public blockchains (for example, Ethereum or Solana) have encouraged a host of innovation from financial services providers, more recently including banks, to consider ways of delivering services on-chain. For example, in June 2025, a subsidiary of Societe Generale, SG Forge, became the first major bank to issue a U.S. dollar-backed stablecoin on the Ethereum blockchain. In addition to stablecoins, banks are actively considering tokenizing deposits, digital representations of a customer’s deposit account, that can be designed to be transferable and collateralized using blockchains. Greater comfort amongst banks with decentralized technologies, alongside regulatory permission to engage with this ecosystem, increases pressure on bank supervisors to also adapt. How they might do so is a question that will demonstrate the capacity of supervision to truly evolve in some of the ways richly discussed by Conti-Brown and Vanatta. Notably, the rise of decentralized infrastructure within banking points to potential gaps in control between what supervisors can recommend that banks do and what banks might actually be able to do in practice. For example, where a bank needs to pay in a blockchain’s native coin (e.g., ETH) to transfer a stablecoin payment from one Ethereum address to another, it ultimately lacks any control over the mechanisms driving the value of that coin and indeed the underlying governance, resiliency, and validation processes of the blockchain infrastructure. How regulators determine the proper delivery of supervisory and advisory functions in this context will need to consider the changing interface between centralized financial services providers and the more decentralized infrastructure on which services may be provided. In other words, even if supervisors try and exercise discretion, there are limits to what they can realistically impose.

Within this more complex and fluid banking industry, regulators are facing calls to include solutions that can more effectively integrate supervisory tools adaptable to decentralization as well as to longer payment supply chains that can include non-bank actors. The deepening sophistication of artificial intelligence, alongside innovations to embed supervision within infrastructure itself, mean that frontline supervisors might be more likely to be a pre-programmed algorithm designed to acquire information, collect its insights and deliver an outcome in the form of a risk rating. This kind of system is arguably a fairly decisive change from the human beings long embedded within the offices of major banks and as described in the book. Automated supervisors can present numerous advantages. For example, they can collect enormous data rapidly from a multiplicity of sources and apply powerful analytics to the task. The need for automation is also necessary, especially as infrastructure itself becomes entirely automated (for example, in the case of blockchains). As such, more artificially intelligent supervisory systems can provide public supervisors with a great deal of information and arguably increase time efficiencies. A key question in designing artificially intelligent bank supervisory systems, however, comes down to questions of discretion. That this discretion is a necessary part of the job is made amply clear by Conti-Brown and Vanatta’s masterwork. It can also be hard for algorithms to accomplish. For example, in the case of trading, studies point to the loss of certain functionalities where floor traders are replaced entirely by electronic trading flows. Here scholars point to the loss of certain kinds of contextual information that human beings tend to glean and that algos easily miss, at times, at a cost to market quality.

Conti-Brown and Vanatta, therefore, while focusing on the historical, point to the future. Its ability to highlight what has worked in the past also hints at how bank overseers need to adapt to ensure that the wins of the past can also be replicated as supervision moves into a brand-new era of digitization, decentralization and disintermediation.

Yesha Yadav is Professor of Law, Milton R. Underwood Chair, and Associate Dean, Culture & Community at Vanderbilt University Law School.