Putting AI to Work in Compliance
What Firms Are Learning About Governance, Efficiency and Human Judgment
AI has quickly moved from an emerging technology to an increasingly important part of the compliance conversation. But as adoption accelerates, a more important question is taking shape: How can firms turn AI’s potential into meaningful, responsible operational impact?
That was the focus of AI in Compliance: From Potential to Operational Impact, the first session in StarCompliance’s (Star) three-part 2026 Global Compliance Benchmark Study Webinar Series.
I was pleased to host the discussion alongside Kelvin Dickenson, Chief Product Officer at Star; Gabby Byrne, Strategic Compliance Lead at NatWest; and Rich Konefal, Chief Compliance Officer at Raymond James Financial. Together, we explored how firms are approaching AI today, the challenges emerging as adoption grows, and what compliance leaders should consider as they move from experimentation toward practical implementation.
The webinar series builds on findings from Star’s inaugural Benchmark Study—download the complete study [Here]—which identified accelerating AI adoption alongside growing concerns around digital assets and prediction markets and increasing pressure to modernize surveillance and information barrier programs.
Here are the six key topics discussed and takeaways from our first session.
1. Governing AI as Adoption Accelerates
As AI becomes more accessible across financial institutions, the conversation is quickly shifting from whether employees should have access, to how that access should be governed.
The panel discussed controlled environments for AI use, access protocols, monitoring, employee training, and how existing surveillance and security frameworks can be extended to AI channels. Protecting material non-public information (MNPI), customer information, and other sensitive data remains central as firms expand their use of these tools.
Governance is also becoming an operational consideration. Firms increasingly need visibility into who is using AI, which tools and models they can access, how they use them, and the associated costs.
Key takeaway: AI governance cannot be static. Controls, monitoring, and policies must evolve alongside both the technology and how employees use it.
2. Building AI Literacy and Critical Thinking
Giving employees access to AI is only the beginning. Getting value from it requires people to understand both its capabilities and its limitations.
A major topic throughout the discussion was the importance of critical thinking. AI can rapidly summarize information, identify patterns, compare data, and support analysis, but its output still needs to be evaluated and challenged.
That makes AI literacy an increasingly important compliance skill. Employees need to understand how to ask better questions, provide the right context, recognize potential inaccuracies, and determine when human expertise and judgment are required.
Key takeaway: AI literacy is about more than learning how to use a tool. Compliance professionals need the knowledge and critical-thinking skills to assess its output and use it responsibly.
3. Using AI to Reduce Manual Work
One of the clearest opportunities discussed was AI’s ability to reduce the amount of time compliance professionals spend finding and assembling information.
For example, when investigating a potential market-abuse alert, an analyst may need to gather historical trading activity, examine patterns and sectors, review previous transactions, and assess the context around an event. Work that could previously take hours can increasingly be assembled in minutes with AI, leaving the analyst to focus on assessing the risk.
The broader opportunity is to use AI to surface the right information at the right time, helping compliance professionals make informed decisions faster without removing humans from the decision-making process.
Key takeaway: The value of AI in compliance may be less about replacing decisions and more about accelerating the work required to make better-informed decisions.
4. Understanding the Other Side of Efficiency
AI can reduce manual effort, but that does not necessarily mean less work.
Our Benchmark Study found that while 23% of respondents reported reduced manual effort from AI, 26% said workloads had increased.
The webinar explored some of the reasons behind that apparent contradiction. Faster reporting can create demand for more reporting. Greater access to insights can raise expectations for faster answers. And if AI increasingly handles routine tasks, employees may spend more time on complex cases that require significant judgment and concentration.
Key takeaway: Firms should measure AI’s impact beyond time saved. Workload, employee experience, skill requirements, and the changing nature of compliance roles also need to be part of the equation.
5. Knowing When to Build and When to Buy
AI has also changed the Build vs. Buy conversation. As tools make it easier to develop internal applications and automate individual workflows, firms have more opportunities to build capabilities themselves.
But the discussion highlighted an important distinction between solving a defined internal problem and developing technology that must operate reliably at scale. More complex compliance systems require ongoing maintenance, integrations, data, regulatory knowledge, and continued development long after the initial solution has been built.
Key takeaway: AI may make building easier, but it does not eliminate the long-term considerations that come with maintaining complex compliance technology.
6. Starting With a Problem, Not AI
For firms still early in their AI journey, one of the most practical themes from the conversation was to resist adopting AI simply for the sake of adopting AI.
Instead, identify a real problem: a repetitive process that takes hours, a workflow dominated by information gathering, or a manual task that creates unnecessary friction. Start there, test how AI can help, learn from the experience, and use those lessons to expand into additional use cases.
The discussion also emphasized that the quality of AI output is closely connected to the quality of the input. Providing sufficient context and asking the right questions can significantly improve the usefulness of the results.
Key takeaway: Successful AI adoption starts with the business problem, not the technology. Find the right use case, demonstrate value, and build from there.
Continue the Benchmark Conversation
The AI discussion was only the first of three sessions exploring findings from the 2026 Global Compliance Benchmark Study.
If you missed AI in Compliance: From Potential to Operational Impact, watch the webinar on demand [HERE] to hear the full discussion on adoption, governance, efficiency, AI literacy, and the changing role of compliance.
The series continues with:
New Markets, Same Risk: Compliance Challenges in Digital and Prediction Markets
Discover how the rise of prediction markets and the evolution of digital asset oversight are transforming employee compliance, and what firms need to do to stay ahead of emerging risks.
Information Barriers Under Pressure: MNPI Governance in a Connected World
Explore the evolving challenges shaping information barriers and MNPI governance, and discover how firms are modernizing control room operations, surveillance programs, and oversight frameworks to keep pace with an increasingly connected risk landscape.
Register today for the remaining webinars and join us as we continue to unpack what the Benchmark Study findings mean for compliance teams in practice.
