Making AI work in regulated environments

By Maciej Nikodemski, Head of Marketing, Inteca

Financial institutions are under growing pressure to modernise their technology stacks while maintaining strict governance, transparency and regulatory compliance. From exchanges and clearing houses to banks and custodians, organisations across the financial ecosystem are exploring how artificial intelligence can improve efficiency, operational resilience and scalability.

Yet integrating AI into highly regulated environments presents a unique challenge. Systems must not only deliver results but also operate within clearly defined processes, maintain full audit trails and ensure that decisions remain explainable and controllable.

This tension between automation and governance is becoming one of the defining questions of the next phase of digital transformation in financial services.

Despite massive investments in artificial intelligence, many organisations are still struggling to translate experimentation into measurable operational outcomes. According to the MIT report “The GenAI Divide – State of AI in Business 2025,” a large majority of companies have yet to see clear, organisation-wide benefits from their AI deployments.

Why does this gap exist?

According to many practitioners, the challenge lies less in the technology itself and

more in the way it is integrated into everyday work. Financial institutions are under growing 

“In many organisations, AI operates next to people rather than with them,” says Habte Woldu, CEO of technology company Inteca. “It may improve the productivity of individual employees, but that doesn’t necessarily translate into higher effectiveness at the organisational level. What’s often missing is consistency, shared context and a structured way for AI to learn within real operational practices.”

From tools to operating models

One way to address this challenge is to rethink the role AI plays in the organisation. Instead of treating it purely as a productivity tool, some companies are beginning to integrate AI directly into their broader operating models.

This approach assumes that AI should become part of the workflow itself rather than an external assistant used in isolated tasks.

One example comes from Inteca, a technology company that began exploring how artificial intelligence could support the entire lifecycle of software development rather than individual activities. Within months, AI systems were assisting across multiple stages of the process - from analysing documentation and proposing architectural solutions to supporting testing and deployment.

According to the company’s internal estimates, productivity in certain stages of the development process increased by up to around 80 percent, while a large share of documentation and code is now created with the support of AI tools.

“The way we build solutions today resembles an intelligent production line,” explains Marcin Parczewski, co-CEO of Inteca. “The process is coordinated from the initial concept through design and architecture to testing and deployment. AI analyses information, proposes solutions and supports execution, while people oversee the process and make the key decisions.”

A changing role for experts

Such a transformation inevitably changes the role of specialists. As AI systems take over many repetitive tasks, human experts increasingly focus on architecture, supervision and decision-making.

Instead of writing every line of code manually, engineers act more like system architects and quality controllers, ensuring that solutions make sense within the broader business and technological context.

“People remain at the centre of the process,” Parczewski emphasises. “AI takes over

repetitive work, but humans are still responsible for understanding the problem, defining the direction and evaluating results.”

For many organisations, this shift represents not a replacement of human expertise but an evolution of it.

From task automation to outcome automation

A key idea behind this emerging model is moving beyond traditional automation.

Historically, automation relied on defining a precise sequence of steps that a system should follow. This works well in stable environments where the same process can be repeated indefinitely.

Knowledge work, however, rarely follows such predictable patterns.

“In complex environments, especially in technology or expert domains, it’s difficult to pre-define every step,” says Woldu. “Instead of automating specific actions, the focus shifts toward automating outcomes - achieving the desired result even if the path to it varies each time.”

In practice, this leads to the emergence of what some researchers describe as “agentic teams” - hybrid teams composed of human professionals and AI agents collaborating on shared tasks and exchanging context and information.

For regulated sectors such as financial services, this approach may prove particularly important. Automation can increase efficiency and speed, but it must also coexist with governance frameworks, operational controls and regulatory expectations.

Models that integrate AI directly into structured workflows, while preserving transparency, repeatability and auditability, may therefore play a key role in the next phase of technology adoption in financial institutions.

The next stage of AI adoption

The broader lesson emerging from early adopters is that successful AI deployment is less about individual tools and more about how organisations redesign the way work is performed. “Using AI is no longer the main challenge,” Woldu concludes. “The real question is how organisations integrate it into their processes, decision-making and collaboration models. Only then can AI start delivering consistent and measurable outcomes.”

In a landscape where many organisations are still experimenting with artificial intelligence without clear operational impact, approaches that combine human expertise, structured workflows and intelligent systems may offer a glimpse of what the next stage of AI adoption will look like.

“In many organisations, AI operates next to people rather than with them.”

Habte Woldu, CEO Inteca

About Inteca

Inteca is a Polish technology company specialising in system architecture, IT integration and digital transformation solutions. The company has been developing approaches that help organisations integrate artificial intelligence into everyday operational and technology workflows.

One example is the PractIQ platform, designed as an AI delivery operating layer that helps enterprise IT teams organise and manage how AI is used across the software delivery lifecycle. The platform supports the standardisation of workflows, introduces governance mechanisms and enables the automation of selected stages of software delivery with the support of artificial intelligence.