ork.AI
In January of 2025, we pointed out in a blog post and demonstrated in a white paper exactly why many banking and financial services organizations—including FinTechs—had already or would soon adopt AI agents to strengthen their AML/KYC operations. We stated, as many others had, that adoption wasn’t a matter of if, but of when. At the time, many banks and FIs were considering the purchase of AI Agents, and some were attempting to build their own AI agents in-house.
Today, the landscape of builders versus buyers of AI Agents has shifted again. WorkFusion counts 10 of the top 20 banks as customers who have decided to buy one or more AI Agents for AML/CFT compliance operations, not build them. And what has happened at the largest banks is happening among FinTechs, regional banks, and non-traditional money handlers. In fact, according to a June 2025 Boston Consulting Group (BCG) analysis, Fintech revenues grew 21% year-over-year in 2024, up from 13% in 2023. Banks are also setting revenue growth targets higher for new products and services than for traditional ones. “Technology resources should be allocated away from cost avoidance initiatives and toward projects that improve new features, enhance integrations, etc.,” said BCG. This means that Engineering and Development teams should not be spending their time on projects focused on cost centers. Instead, they need to help deliver revenue-driving new products and services.
Self-Builds Don’t Make Sense
If you work in Engineering or Development at a bank or FinTech, you recognize that self-builds of AI Agents are very lengthy, high-cost propositions that, even if technically viable, often fail. Moreover, if AI Agents already exist for cost center operations, then a self-build makes even less sense. And when it comes to compliance operations, it’s a massive ask to expect a self-built solution to be a regulatory-ready, end-to-end solution that can solve customer satisfaction, revenue impact, and compliance staffing challenges.
But rather than discuss the doom and gloom of self-builds, let’s flip the script to the brighter side and discuss how buying AI Agents not only delivers greater and faster ROI, but it also sets the stage for geometric value growth across the organization. And when we say “geometric,” we’re actually going to demonstrate that versus the mere “multiplicative” gains targeted (and rarely delivered) by self-builds of AI Agents.
Here’s a quick brush-up if you’ve not kept up with all those math classes: A multiplicative progression—expressed as (2,4,6,8,10…)—is a far less lofty accomplishment than a geometric progression, which can be expressed as (2,4,8,16,32…). See how you get to 32 after just five stages of geometric progression? That’s far and away superior to the ‘10’ realized after five stages of multiplicative progression.
How WorkFusion AI Agents deliver geometric value growth
With just a single pre-built WorkFusion AI Agent in place, a FinCrime compliance organization has the key foundational element needed to speed value growth in terms of:
- Compliance efficiency
- Work scaling
- Information accuracy
- Informed escalations/investigations
- Consistent reporting
- AI explainability to regulators.
Now, consider what happens when your internal Engineering/Development team wants to build a solution themselves. They must scale their internal development work, and that requires significant effort navigating complex business units and adhering to a variety of compliance requirements across an expansive portfolio of offerings. This holds true even for engineering-first organizations.
By contrast, here’s the stage-setting for geometric value growth at a WorkFusion customer—a Top 25 US Bank:
- Without increasing compliance headcount, the customer launched new, revenue-generating payment products while maintaining full adherence to regulatory requirements.
- Deploying multiple AI Agents across business lines has given the customer organization a standardized, efficient approach to addressing operational inefficiencies.
Another WorkFusion customer—a Top 10 US Investment Bank—was already using WorkFusion’s AI Agent named ‘Tara’ to review alerts in their payment sanctions screening process. Soon, recognizing Tara’s capabilities, the bank’s leaders decided to launch a Banking-as-a-Service offering that required scalable, effective payment screening to meet compliance standards. By redeploying the existing AI Agent, Tara, the bank accelerated compliance operations for the new business line—streamlining risk controls without adding manual overhead.
Suddenly, the bank new business line was processing alerts at a rate of 1.5 million per year with an error rate of just 0.03%, of which 0% were unexplainable. That’s a massive business transformation backed by an already-built and already-in-use AI Agent. To ask an internal development team to do that, and deliver such high ROI, would be unrealistic. The only reason we don’t call that value growth exponential is that exponential growth happens at a steady rate. Value growth with our pre-built AI Agents can sometimes be sudden, and that burst of value makes it geometric.
In the end, the enjoyment you will have by measuring the geometric value gains delivered by WorkFusion AI Agents will only be rivaled by the levels of frustration you will have saved your Engineering and Development teams in a self-build.