AI in Finance Is Mainstream, But the Operating Model Is Still Legacy
Artificial intelligence governance in finance refers to the systems, rules, people and processes that ensure AI tools used in financial planning and decision-making are reliable, compliant, explainable and aligned with long-term business goals, rather than being isolated pilots or unchecked algorithms that operate outside traditional risk and control frameworks. Artificial intelligence in finance has hit an inflection point: 76 per cent of organisations already use AI in financial planning. That headline figure looks like success, but it hides a structural problem. Adoption is racing ahead while the operating model that surrounds AI remains stubbornly traditional. According to KPMG, "competitive advantage will increasingly belong to organizations that align their finance operating model around AI" and sustain that alignment as AI becomes more capable and autonomous. The uncomfortable truth is that most finance teams are treating AI as a powerful add-on, not as the organising principle of how decisions are made. As long as AI sits at the edge of the operating model, firms will get short-term gains but miss the deeper transformation.
Governance: The Biggest Risk Blind Spot in Financial Services AI
For financial services AI risk, governance is not a paperwork exercise; it is the difference between scalable advantage and regulatory backlash. The report states plainly that governance is the biggest challenge to realising AI’s value in finance. Firms are deploying models faster than they can build rules on data quality, model validation, auditability and accountability. This is a paradox: finance is one of the most regulated domains, yet AI governance in finance is lagging the technology itself. In a sector where regulatory scrutiny and professional judgement shape every output, weak oversight turns AI from an asset into a liability. Some organisations recognise this and are increasing human-in-the-loop oversight, with 33 per cent doing so as a direct response to AI concerns. That is a start, but bolting human review onto poorly governed systems is a stopgap. Without clear ownership, escalation paths and model lifecycle controls, AI adoption challenges will compound, and risk officers will be left cleaning up after algorithms they do not fully understand.
Talent: Training Users Without Redesigning the Workforce
The talent picture shows why governance frameworks struggle to move beyond theory. Many organisations are upskilling existing staff, but only 28 per cent are reassessing the talent profiles needed for an AI-driven future. In other words, firms are teaching people to use tools without rethinking who they need in the first place. This creates a widening gap between tool proficiency and true AI literacy. A total workforce approach is rare: while 38 per cent are upskilling finance and internal audit teams for AI-enabled processes, only 28 per cent are hiring people with new skill sets. The message is clear—most firms are stretching current roles instead of redesigning them. That is risky. Effective AI operating models demand hybrid profiles: professionals who blend financial expertise, data fluency and model risk understanding. Without them, AI governance finance initiatives become box-ticking exercises managed by people who either lack technical depth or cannot connect it to business outcomes. Talent shortages are not a side issue; they are a core driver of AI adoption challenges.
Data Fluency and AI-Centric Operating Models: Where the Real Edge Lies
Even with high adoption, many finance teams do not know where to start in turning AI into sustained advantage. The report notes that 36 per cent of organisations see improving data quality, integration and system interoperability as their biggest opportunity to extract more value from AI in finance. That is telling: the limiting factor is not the algorithm, but the ability to feed it reliable data and interpret the output. Data fluency is described as the most critical capability—assessing data quality, interpreting AI outputs and explaining findings in terms the business can act on. This is the backbone of AI-centric operating models. When finance teams own data fluency and embed AI into core planning, forecasting and performance management processes, AI stops being a bolt-on and becomes the default way decisions are made. The firms that will win are those that treat operating model redesign as the main product of AI adoption, not a side project. Technology is easy to buy; rewiring processes, responsibilities and data flows around AI is where competitive edge is earned.
Conclusion: High AI Adoption Without High Maturity Is a Strategic Trap
The headline number—76 per cent of organisations using AI in financial planning—signals that AI is now standard kit in finance. But high adoption with low maturity is a strategic trap. Firms that rush ahead without investing in governance frameworks, AI operating models and redesigned talent will face mounting financial services AI risk, from model errors to regulatory challenges. The evidence is consistent: governance is cited as the biggest challenge, human oversight is being increased reactively, and only a minority of organisations are rethinking their workforce and data capabilities. That is not a sustainable position. The next phase of competition will not be about who uses AI, but about who can make AI safe, explainable and central to how finance operates. The conclusion is blunt. Deploying AI is now table stakes. Treating AI governance, talent strategy and operating model redesign as core leadership priorities is what will separate tomorrow’s winners from the firms that merely experimented.
