The finance industry's AI investment is surging in 2026, with 35.7% of financial services professionals naming it their top tech priority, outpacing cloud and data projects, according to Alithya's 2026 financial services survey. Yet many firms struggle to pinpoint where value lies, even as AI shifts from answering questions to moving money and making decisions on behalf of customers. A new analysis blending the output of two leading AI models-Claude and ChatGPT-highlights five priorities that financial institutions should tackle immediately.
Proving ROI and killing what doesn't work
The top priority is establishing clear return on investment and being willing to stop initiatives that fail to deliver. As organizations push to prove returns in AI for Finance, the pressure to cut underperforming projects is mounting. AI projects require nimble teams and honest assessments, with senior leaders ready to pull the plug when doubt arises. "Things will not go to plan," the editorial said, "but rather than letting projects run to a conclusion, it's imperative that if there is doubt, the plug gets pulled and everyone moves on."
Preparing for agentic finance and governance
AI is moving from answering questions to performing actions-moving money, changing investments, negotiating bills, or managing debt. The FCA's Mills Review found that about 11 million UK adults are likely to use autonomous AI within pre-set goals, though trust and control remain concerns. The emerging model is human-on-the-loop, where people check and approve AI's work rather than building from scratch. This requires governance built into design from the start, with secure guardrails, strong monitoring, and clear accountability for mistakes.
Fixing data and controlling concentration risk
Data quality remains a stubborn barrier. KPMG's 2026 Global AI in Finance report shows that 36% of organizations say improving data quality, integration, and system compatibility is their best route to more AI value. Many large financial firms, particularly in the UK, operate with a patchwork of platforms and data models inherited from acquisitions, and the experts who built those legacy systems are retiring. Meanwhile, reliance on the same foundation models, cloud platforms, and data providers creates concentration risk-a single failure could cascade across institutions. Firms need portable systems, alternative providers, and contingency plans for key supplier failures.
AI for financial wellbeing, not just cost cutting
Most firms have started with productivity gains-summarizing documents, writing code, automating service. The bigger opportunity is helping people understand money, deal with debt sooner, and make better decisions without judgment. AI can create personalized, distinctive customer touchpoints, but the industry tends to value extraction over value creation. The editorial posed a direct question: "Once the machines have perfected the extraction, what will be left to set you apart, except the value you create?"
Why this matters for finance professionals
For those in finance roles, the message is clear: the AI investment wave demands discipline. Leaders must be ready to kill projects that don't show results, build governance into agentic systems from day one, and tackle data messiness before it undermines everything else. The competitive edge won't come from efficiency alone-it will come from using AI to genuinely improve customers' financial lives.
Your membership also unlocks:
