From Compliance to Connection: Rethinking Financial Services

Money20/20 in Las Vegas closed out a tumultuous year. Digital experiences have constantly evolved. Customer behaviour remains unsettled. There’s no playbook. Across the floor and stage, I felt both anxiety and opportunity. Most panels centred on stablecoins and AI, but the excitement came from experimenting with technology that scales rapidly while making a real difference for customers. 

Key Takeaways 

1. Too much compliance, not enough customer

Many live use cases focused on regulatory compliance rather than enhancing customer experience. Teams are leaning on AI to adapt to shifting regulations and streamline fraud/risk workflows. Although that’s useful, it’s not the leap forward in customer experience the TradFi world spends billions trying to deliver. Agentic systems were the buzz, yet many demos still imagined AI engagement categories like “personalized recommendations” or “onboarding” rather than specific use cases that deepen loyalty or drive acquisition.

2. Payments’ next act is UX, not rails

Cross-border payments were a hot topic, indicating that liquidity creation is now a customer differentiator. Stablecoins were positioned as the 'Uber vs. taxis' moment for settlement by leaders from Anchorage Digital and Western Union, raising the question: What’s the end-to-end UX that makes a traveller or small business feel the benefit? Avoiding a negative experience (declines, delays, fees) may not be enough. The winners will create positive moments that customers share.

3. Tool sprawl is real; time to think like an OS

Dashboards often mask unstructured data. No-code tools may create the appearance of integration. Still, the real opportunity lies in composable operating systems: interfaces tailored for employees that allow AI to act across systems on structured data foundations.

4. Adoption beats architecture

Here’s a particularly memorable line from Greg Ulrich, chief AI and data officer at Mastercard: “It’s not the technology. It’s how you train your employees to adopt it.” FinServ employees often regard AI like a blank page: It has endless possibilities but no clear starting point. Tried-and-true tools like Excel become the competitor, not other LLMs. The industry needs a tailored first-time-user experience for internal teams, manufactured scarcity (get value fast or access goes away) and psychological safety so employees can learn in public. 'Vendor training' alone doesn’t move the needle. Organisations want embedded enablement that maps to their workflows and KPIs.

5. Brand in the automation age

Digital-first innovators like Chime are creating 'golden datasets' and brand brains trained on greatest hits. That accelerates production, but it can also threaten evolution if not done carefully. If agents standardise execution and everyone has similar frameworks, brand tone becomes the primary differentiator. Placing humans in the loop for empathy, taste and boundary-pushing creativity, then letting AI scale what resonates, acts as the counterweight toward commoditization.

6. To go fast, go broad

Banks often run parallel vendor tracks before consolidating once concepts prove out. For agencies and systems integrators, there’s a role to play as a consultative layer in RFPs, helping banks and fintechs pressure-test integration, governance and change-management plans before code is ever written.

Where we go from here:​

• ​Define agentic jobs-to-be-done. If an agent can recommend a product, can it also build loyalty? What are the emotionally aware use cases that a third-party agent can’t perform as well as our owned channels?
  
• Design for desire. Motivation is the key predictor of adoption. Internal first-time user experiences should look like our best consumer onboarding.
  
• Own the integration interface. If we want to be indispensable, we need a composable layer that tames unstructured data and lets AI act across systems with governance.
  
• Measure employee experience. What’s the “employer experience” score for internal tools? If our agents level the playing field, how do we help in-house teams differentiate above the glass (brand, tone, choreography) while hardening below the glass (data, controls)?
  
• Shift from trust to growth. Trust gives us the right to play; growth comes from personalised, joyous moments, especially during critical money moments like income deposits, travel and family milestones.
  

If 2025 was about proofs of concept, 2026 will centre on scaling frameworks into meaningful experiences for customers and employees. We can’t let this all innovation merely lead to a better chatbot. Our responsibility lies in making the whole system, from DeFi to TradFi, feel personal, safe and surprisingly helpful.