By Tim Fleury, Chief of Staff, DCM Services
As artificial intelligence and automation become more common in highly regulated industries, the challenge for business leaders is shifting their approach while staying compliant. The question is no longer whether these organizations should adopt AI, but where and how it should be used.
The organizations that gain the greatest advantage won't necessarily be the ones that automate the most. They will be the ones that exercise the best judgment about when it’s warranted, or even responsible to use AI, when human oversight is necessary and when people should remain at the center of the customer experience.
That distinction matters in regulated businesses, where efficiency is only one measure of success. Compliance, information security, customer experience and trust matter, too.
The question for leaders is no longer simply what AI can automate. It is what AI should automate.
Where Should Companies Use AI and Automation?
One of the easiest mistakes organizations can make is starting with the technology and then looking for places to deploy it.
A better approach is to start with the work. Asking questions like these give a much deeper and more valuable insight into opportunities for AI to supplement your workforce:
Where are employees spending time on repetitive, administrative activities?
Where could automation improve consistency?
Where could AI free people to focus on higher-value work?
Where does the work require context, empathy or judgment that technology cannot adequately replicate?
Asking yourself those questions help establish a natural division between tasks that can be automated, work that can be augmented by technology and responsibilities that should remain human.
We've seen that distinction firsthand at DCM Services. Many of the interactions our teams have involve people navigating the loss of a loved one. That makes it especially important to be deliberate about where technology fits.
We use AI to support defined administrative activities such as call preparation and documentation, helping create more consistent and complete records while allowing representatives to focus more fully on the conversation. AI also supports quality review, enabling us to evaluate more interactions consistently.
The objective isn't to remove the person from the process. It's to make that person more effective.
In that sense, one of the most valuable applications of AI may be surprisingly human: using technology to create more room for people to be people.
Why Is Human Judgment Still Essential?
AI is powerful because it can process information, recognize patterns and perform well-defined tasks quickly and consistently. But consistency and judgment aren't the same thing.
That distinction becomes especially important when customer interactions are sensitive or complex.
A model can recognize patterns in language. A skilled representative can hear uncertainty or emotion in someone's voice and adjust the conversation. They can recognize when slowing down matters more than completing an interaction quickly and can apply context that may never appear explicitly in the data.
At DCM Services, we hire and train specifically for those human capabilities because empathy and listening are central to the work our representatives do. AI can support that work, but it isn't a substitute for the judgment required to do it well.
The same principle applies more broadly across regulated industries. Businesses routinely encounter situations that are ambiguous, sensitive or context dependent. A process may have clearly defined rules while still requiring someone to determine how those rules apply to a particular situation.
Human oversight is what determines when automation is appropriate rather than simply being a signal of there being a failure of automation.
As AI becomes more capable, leaders will increasingly have to distinguish between whether a machine can perform a task and whether it should perform that task independently.
How Should Regulated Companies Govern AI?
In regulated environments, AI governance cannot be separated from AI innovation.
Once technology has been deployed, it’s too late to worry about data privacy, information security, regulatory obligations, client requirements and internal controls. Those issues have to be considered from the beginning before you commit the time, money, and energy into adopting new technologies or automations.
There is sometimes an assumption that governance slows innovation. I see it differently.
Strong governance is what makes sustainable innovation possible.
A compelling demonstration of what an AI tool can do isn't enough to establish that it is appropriate for a business environment. The technology must perform under real operating conditions. Output needs to be evaluated for relevance, efficacy, and compliance. Risks have to be understood. Appropriate human review procedures should not only exist but be settled upon before rollout. And sensitive information must be protected.
Our experience at DCM Services has reinforced the importance of that discipline. We operate in an environment where compliance, data security and quality control are already embedded in how we do business, and we believe AI should be held to that same standard. We start with defined use cases, pilot them in a controlled way, evaluate the quality and accuracy of the output, and maintain human review before relying on that output operationally. Just as importantly, data privacy is a design consideration from the outset: customer data isn't used to train models, and AI applications must operate within our existing security and compliance framework.
For us, responsible AI isn't a separate governance exercise layered onto the technology after the fact. It's an extension of the controls, risk management and accountability already expected in a highly regulated business. That discipline gives us the ability to explore where AI can create value without lowering the standards our clients and their customers expect. The business case for automation therefore can't be evaluated solely on operating efficiency.
A technology that reduces operating costs while introducing unacceptable compliance, customer or reputational risk hasn't eliminated costs. It has simply shifted them.
In regulated industries, trust has to be part of the return on investment.
When is an AI Use Case Ready to Scale?
The speed of AI development creates understandable pressure to move quickly from experimentation to enterprise-wide adoption. A disciplined implementation begins with a clearly defined business problem and a controlled use case. Organizations can then evaluate accuracy, security, operational impact, compliance implications and the experience of the employees and customers affected before expanding it.
Leaders should ask a few straightforward questions:
Does this solve a meaningful business problem?
Does it improve the experience for employees, customers or clients?
Can it operate within our security, privacy and compliance requirements?
Do we understand where human judgment is still necessary?
Can we demonstrate that it is producing a better outcome?
These questions create a more durable framework for AI adoption than pursuing the newest capability simply because it is available. Technology will continue to change. Models will become more capable. Economics will evolve. New use cases will emerge. That makes the framework an organization uses to make decisions about AI just as important as the technology itself.
Why Judgment Will Become a Competitive Advantage
As AI becomes more accessible, many of the capabilities that appear differentiating today will eventually become widely available. Organizations will increasingly have access to similar models, tools and automation platforms. Simply having AI is unlikely to create a durable competitive advantage.
The differentiator will be how organizations choose to use it.
Companies that understand their customers well enough to recognize where automation improves an experience and where it diminishes one. Responsibility rests with the leaders who recognize that technological capability and business wisdom are not the same thing.
For highly regulated organizations, this doesn't require choosing between innovation and compliance or between automation and human connection. The opportunity is to design systems in which each reinforces the other. AI can improve consistency, reduce administrative work and help organizations evaluate information at a scale people alone cannot. It can also give employees more time to focus on the work requiring their expertise, empathy and judgment. But realizing those benefits requires knowing where the boundaries should be.
As AI becomes ubiquitous, access to technology will become less differentiating. The judgment surrounding its use will become more differentiating.
The organizations that lead in the age of AI won't simply be those with access to the best technology.
They will be the ones with the judgment to know how and when to use it.
