Money20/20 Europe recap: the AI conversation turns serious

Money20/20 Europe 2026 made one thing clear. AI adoption is not slowing down, but trust is still catching up.

The industry is not debating whether AI is useful anymore. It is grappling with how far it should be trusted, especially when money, regulation, and customer outcomes are directly involved.

Agentic AI was everywhere on stage and across the show floor. But the most important signal was not excitement about autonomy. It was the caution, maturity, and humanity surrounding it. 

Across keynotes, executive conversations, and vendor discussions, three themes stood out.

1. AI Is Moving From Assistant to Operator

For the past few years, most organizations have used AI to help employees work faster. Increasingly, however, financial institutions are exploring AI agents that can perform work on behalf of humans.

The concept of agentic AI was everywhere at Money20/20. But while enthusiasm is high, so are the stakes.

Unlike marketing content or internal productivity workflows, financial services operate in an environment where AI decisions can directly impact people’s money. Whether it’s fraud detection, lending decisions, investment recommendations, or payments, institutions must determine how much authority they’re willing to give autonomous systems.

Forrester’s post-event recap explicitly says agentic AI was one of the three defining themes of the conference, but the conclusion wasn’t “let the agents run everything.” The analyst’s takeaway was that as AI becomes an “economic actor,” trust becomes a system-level requirement and firms must build “trust-first operating models.”

2. Trust and Governance Are Becoming Competitive Advantages

Perhaps the most notable shift from previous years was how often governance appeared in AI conversations.

Financial institutions are under pressure to move faster. Real-time payments, digital banking experiences, and increasing customer expectations are pushing organizations to accelerate software delivery.

At the same time, regulators, security teams, and risk leaders are asking important questions:

• How do we audit AI-driven decisions?
• Who is accountable when AI makes a mistake?
• How do we maintain compliance as systems become more autonomous?
• What controls are required before AI can be trusted with customer-facing decisions?

The broader industry is arriving at the same conclusion.

FinTech Futures notes that although one of the event’s four major content pillars was “AI and the Agentic Age,” much of the discussion centered on governance, compliance, and regulatory readiness. One featured panelist argued that agentic commerce requires compliance controls to move “upstream” in the process, while an FCA innovation leader remarked that the governance challenge has arrived sooner than expected. 

This mirrors what we’re seeing among engineering leaders. AI adoption is accelerating, but so is the realization that governance models, operating structures, and workforce capabilities must evolve alongside it.

It’s a challenge we’ve written about previously in the context of AI-driven software development and contractor governance. As AI expands the speed and scale of software creation, organizations need new mechanisms to ensure quality, accountability, and oversight.

3. AI Is Everywhere. Humans Still Matter.

One of the more interesting observations from the show floor was that while nearly every company had an AI story, many of the most thoughtful conversations focused on the continued importance of humans.

Industry observers walking the exhibit hall noted that AI had become nearly ubiquitous among vendors and exhibitors. Yet many organizations emphasized human oversight, transparency, and accountability as essential ingredients for successful deployment.

That reflects what we continue to hear from engineering leaders.

The future isn’t human versus AI.

It’s human + AI.

The highest-performing organizations aren’t replacing people with AI. They’re redesigning workflows so humans and AI complement each other’s strengths.

AI can accelerate analysis, automate repetitive work, surface recommendations, and increase productivity.

Humans provide judgment, context, accountability, creativity, and trust.

That combination is particularly important in financial services, where customers, regulators, and stakeholders need confidence in the systems making decisions on their behalf.

A New Question Is Emerging: How Do We Measure Success?

One of the more memorable conversations we had at the Karat booth involved a question we hadn’t heard a year ago:

“Can you assess how efficiently an engineer burns through AI credits?”

The question may sound simple, but it points to a much larger shift.

Organizations are moving beyond Tokenmaxxing as the main measure of AI adoption. They’re beginning to ask how effectively teams use AI.

The way we’re thinking about this question in the context of NextGen interviews is about assessing cost-efficient AI use. 

Organizations are looking for ways to produce working code for less spend (whether by elevating contractor performance, consolidating IT service providers, streamlining workflows, or increasingly by not burning a frontier model and a maxed-out context on trivial edits, no wasted runs). 

In practice, this looks like evaluating an engineer on things like selecting the most appropriate model for a given task, proper context scoping, and how they approach intentional cost/quality trade-offs. 

The challenge isn’t simply deploying AI tools. It’s building teams capable of using them responsibly and effectively.

Help Shape Our Next AI Workforce Study

Throughout Money20/20 Europe, we heard leaders wrestling with many of the same questions:

• How should organizations govern AI-generated software?
• What skills will define high-performing engineers in an AI-enabled world?
• How do we measure AI productivity without encouraging the wrong behaviors?
• Where should humans remain in the loop?
• What does accountability look like when AI agents become part of everyday workflows?

These are exactly the kinds of topics we’re exploring in Karat’s upcoming AI Workforce Transformation research.

We’re currently gathering input from engineering and technology leaders on the questions that matter most.

If there’s a question you think the industry should be asking, we’d love to hear it.

Submit your ideas for our next AI Workforce Transformation Survey here: https://karat.com/ai-software-engineering-survey/

We’ll use the responses to help shape future research on how AI is transforming software engineering, hiring, workforce development, and organizational performance.