A paradox sits at the heart of modern banking's relationship with artificial intelligence: nearly every executive, analyst, and technologist in the industry agrees that AI is transformative, yet the sector's adoption curve remains uneven, its governance frameworks incomplete, and its promise only partially realised. That tension — broad consensus without commensurate action — is precisely what financial commentator and author Chris Skinner has begun to dissect, applying the classic SWOT (Strengths, Weaknesses, Opportunities, Threats) lens to one of the most consequential technological shifts the industry has ever faced.
Writing on his long-running publication The Finanser, Skinner frames what he calls a "wonderful problem" with AI in banking: the near-universal agreement on its importance has not translated into uniform clarity about how institutions should harness it, govern it, or even fear it. The observation is pointed. When an industry reaches consensus that a technology matters without yet agreeing on what to do about it, the result is often a dangerous combination of over-investment in superficial applications and under-investment in foundational infrastructure.
Skinner's current intellectual preoccupation centres on three interconnected concepts: Agentic AI, autonomous commerce, and programmable money. Each represents a distinct but reinforcing layer of a future financial system that operates with far greater automation, self-direction, and embedded intelligence than today's institutions can support. Agentic AI — systems capable of independently planning, reasoning, and executing multi-step tasks without continuous human prompting — is perhaps the most disruptive of the three. Unlike the large language model chatbots that captured public imagination in the early 2020s, Agentic AI does not merely answer questions; it acts on behalf of its principals, making it a fundamentally different risk proposition for regulated financial entities.
Autonomous commerce, the second pillar of Skinner's framework, extends that logic into the transactional layer. When AI agents can negotiate contracts, execute trades, initiate payments, and manage counterparty relationships without moment-to-moment human oversight, the traditional compliance and audit architecture of banking faces serious structural stress. Know Your Customer (KYC) and Anti-Money Laundering (AML) frameworks designed for human-initiated transactions must be reconceived for a world where the initiating party may itself be a software agent acting on behalf of a customer, a fund, or another machine.
Programmable money — whether in the form of central bank digital currencies (CBDCs), tokenised deposits, or smart-contract-governed stablecoins — completes the triangle. It is the substrate on which Agentic AI and autonomous commerce ultimately run. Money that carries embedded conditions, that can settle instantaneously against programmable triggers, and that requires no correspondent banking chain to cross borders, creates both extraordinary efficiency gains and equally extraordinary systemic risk if the underlying logic contains errors or is exploited. Regulators at institutions such as the Bank for International Settlements and the European Central Bank have published extensively on the governance challenges programmable money introduces, but policy frameworks remain nascent relative to the pace of technological development.
Skinner is codifying this analysis into a forthcoming book titled Birth of a Unicorn, the latest volume in his "Pulp Finction" series, scheduled for release in December. The project signals that the conversation around AI in banking has matured beyond conference keynotes and proof-of-concept press releases into territory worthy of structured, book-length treatment. That Skinner — whose prior work has chronicled the rise of digital banking, open finance, and fintech disruption — is now directing his attention so squarely at Agentic AI, autonomous commerce, and programmable money is itself an indicator of where the industry's most serious strategic questions now lie.
The SWOT framework, while familiar to any MBA graduate, carries particular analytical power when applied to a technology still in formation. On the strengths side, AI's capacity for real-time fraud detection, hyper-personalised credit underwriting, and operational cost reduction is already demonstrable across tier-one institutions. The weaknesses, however, are equally concrete: model opacity, data quality dependencies, regulatory uncertainty, and the persistent challenge of integrating AI into legacy core banking infrastructure built on decades-old architecture. The opportunities are vast — from democratising access to sophisticated financial advice to enabling entirely new product categories — while the threats range from adversarial AI attacks on financial systems to the competitive pressure exerted by technology companies that face lighter regulatory burdens than incumbent banks.
What This Means for the Industry
The significance of Skinner's framing is not merely academic. Financial institutions that treat AI as a productivity tool — a faster search engine or a more efficient back-office processor — will find themselves structurally disadvantaged against competitors and new entrants that treat it as an architectural principle. The shift toward Agentic AI means that banks must begin designing their operating models, risk frameworks, and customer contracts around the assumption that many future financial interactions will be machine-to-machine, governed by programmable money rails, and executed without a human decision point in the critical path. Institutions that wait for regulatory certainty before building that capability may find the window to compete has already closed. The wonderful problem Skinner identifies is, in the end, not so wonderful for those who mistake consensus for readiness.
Written by the editorial team — independent journalism powered by Codego Press.