Global financial regulators have moved decisively to shape the rapid spread of artificial intelligence across capital markets. The International Organization of Securities Commissions has released a new Supervisory Toolkit for AI Use in Capital Markets, giving regulators around the world a structured methodology for overseeing AI-based systems. The publication lands at a pivotal moment: fresh industry data shows that finance sector AI adoption has more than doubled in just three years, and the governance infrastructure is only now beginning to catch up.
The IOSCO toolkit covers governance standards, disclosure requirements, recordkeeping obligations, and audit trail expectations. It sets clear expectations around investor protection and market integrity, areas where AI failures do not stay contained. A flawed model in a trading system or a biased credit-scoring algorithm does not just hurt one institution. It can ripple through interconnected markets and affect millions of consumers who never knowingly interacted with the technology behind the decision.
Alongside the IOSCO guidance, banking regulators in both the United States and Europe have moved to require AI explainability as a compliance baseline rather than a best practice. If an AI system denies a loan application or flags a transaction as potentially fraudulent, the institution must now document exactly why, tracing the logic back through the model with a complete audit trail. For banks that deployed early black-box models without auditability in mind, that requirement is prompting expensive and time-consuming retrofits.
The 2026 Global AI in Financial Services Report from the Cambridge Centre for Alternative Finance paints a complex picture of where the industry actually stands. Fifty-six percent of finance leaders now use AI in some capacity, double the adoption rate recorded in 2023. But beneath that headline, the picture is more uneven. Only 17 percent of finance teams are embedding AI into their core workflows in a meaningful way. Nearly half, 45 percent, remain in what the report calls limited pilot mode: evaluating tools, running proofs of concept, but stopping short of full commitment.
The gap between fintechs and incumbent banks has widened noticeably. Fintechs report advanced AI adoption rates of 47 percent, compared to 30 percent for traditional institutions. That 17-point lead reflects structural advantages that are difficult for legacy players to overcome quickly: leaner operations, more modern technology stacks, and organizational cultures built around rapid iteration. For established banks, the challenge is not locating AI tools. It is integrating them into systems built over decades before machine learning was a practical reality.
One area where adoption is surprisingly uniform across institution types is agentic AI. These are systems capable of taking autonomous, multi-step actions rather than responding to individual prompts one at a time. Think of an AI that does not just flag a suspicious transaction but investigates it, cross-references other accounts, and generates a compliance report without a human in the loop for each step. The CCAF report found that 52 percent of industry respondents are already actively adopting agentic AI, a striking figure for technology that was largely theoretical just two years ago. Gartner has forecast that 40 percent of all business software will include AI capable of completing end-to-end tasks independently by the close of 2026.
The Data and AI Summit held in San Francisco in mid-June brought together AI leaders from Morgan Stanley, JPMorganChase, and Mastercard to discuss what many described as a pivot from experimentation to infrastructure. The recurring theme was that 2025 was the year of proving that AI could work in finance; 2026 is the year of making it reliable enough to stake the business on.
CFOs are feeling pressure from both directions simultaneously. Board-level expectations for efficiency gains driven by AI are rising steeply. At the same time, compliance teams are flagging regulatory risk from systems rushed into production without adequate documentation. The CFO Connect State of AI in Finance 2026 report found many finance leaders caught between the two: aware that moving too slowly cedes ground to competitors, and equally aware that moving too fast invites scrutiny from regulators who now have both the IOSCO toolkit and the political mandate to use it.
For fintechs with strong governance foundations already in place, the regulatory attention may function as a competitive moat rather than a burden. Firms that invested early in explainable AI and audit infrastructure are better positioned to clear the bar that regulators are setting, while less-prepared incumbents scramble to retrofit documentation onto systems that were never designed with supervision in mind.
The direction is unmistakable. AI in financial services has crossed a threshold. It is no longer a peripheral capability tested in innovation labs. It is becoming load-bearing infrastructure, and regulators worldwide are racing to ensure that the foundations are solid enough before the weight becomes too great to safely manage. Keep up with the latest developments across AI and technology at Mylistingo.
