Beyond the Hype: Realizing AI’s Potential in Compliance

Lucinity’s “Intelligence in Action” series recently hosted an engaging webinar, led by Guðmundur Kristjánsson (GK), Founder and CEO of Lucinity. The “The Things You're Not Thinking About When Implementing AI in Compliance” session brought together two experienced professionals from Oliver Wyman—David Choi, Partner, and David Carretero, Principal.

This discussion examined the transition from AI’s initial excitement to its practical applications in compliance. It focused on practical strategies for addressing adoption challenges, identifying specific use cases, and applying AI effectively to deliver significant value for compliance teams. Key questions tackled included:

• How can compliance teams identify high-value AI use cases?
• What steps are essential for ensuring successful implementation and change management?
• How do we define and measure ROI in compliance-focused AI initiatives?

This recap distills the highlights, offering practical insights to help organizations use AI effectively while remaining resilient to industry challenges.

Key Insights from the Webinar

The webinar brought together three leaders in compliance and AI to discuss the progression from AI’s possibilities to practical implementation. Hosted by Guðmundur Kristjánsson (GK), Founder and CEO of Lucinity, the discussion featured David Choi and David Carretero from Oliver Wyman. Their combined expertise highlights how compliance teams can address challenges, effectively use AI, and achieve the right balance between automation and human judgment.

Introduction

GK began by addressing a key challenge for compliance leaders: transitioning from theoretical AI possibilities to practical implementation. Noting that 60% of compliance leaders find it difficult to define actionable AI use cases, he centered the session on ways to overcome these challenges. He emphasized that implementing AI isn’t about installing new technology alone—it involves rethinking processes, identifying core problems, and preparing for unavoidable obstacles.

With insights from his guests, GK promised a candid discussion on AI’s impact on compliance processes, the importance of change management, and building confidence for future-ready compliance teams.

Q: How has the AI hype cycle influenced its adoption in compliance?

GK began by asking Choi to share his thoughts on AI's progression through its hype cycle and its present influence on compliance.

Choi explained that while the initial excitement around AI has subsided, this quieter phase marks a key point where organizations focus on pilots and proof of concepts (POCs). Generative AI adoption has been cautious, largely due to concerns about governance and security. At the same time, machine learning applications have seen notable advancements, especially in areas like transaction monitoring.

Firms still relying on rules-based systems face high false-positive rates, prompting many to explore machine learning solutions. This change is gaining speed as organizations transition from enhancing rules-based models with machine learning to completely replacing outdated systems. According to Choi, the next 12–18 months will likely see broader AI deployments, as POCs pave the way for production-scale implementations.

Q: What are the first steps organizations should take to embrace AI in compliance?

In response to GK’s inquiry about actionable steps, Choi emphasized starting with strategy. Organizations must first identify areas where AI can deliver tangible benefits, such as reducing compliance costs or enhancing detection accuracy. He underscored the importance of building robust business cases that outline measurable ROI.

Carretero added to the discussion by highlighting the need to democratize AI within organizations. Encouraging employees to experiment with tools like Microsoft Co-Pilot allows teams to explore AI’s capabilities, generate excitement, and identify potential use cases organically. This bottom-up approach complements top-down strategies by ensuring buy-in from all levels of the organization.

He also stressed the importance of collaboration between compliance, business, and IT teams. For complex use cases, cross-departmental alignment is important to ensure both the technical feasibility and strategic relevance of AI deployments.

Q: What should compliance leaders know about transitioning AI from pilot stages to production?

GK asked how companies can move beyond pilots to implement AI at scale effectively. Choi emphasized that pilots and proofs of concept (POCs) are foundational to ensuring robust deployments. Organizations frequently begin with small-scale experiments but face difficulties expanding to larger production due to fragmented systems and unclear objectives. Choi suggested creating a detailed roadmap with clear metrics to scale successful pilots while tackling governance, security, and resource alignment from the outset.

Carretero added that firms often underestimate how important cross-functional collaboration is during this phase. Compliance teams must work closely with IT, data science, and operations to integrate AI into existing workflows while identifying areas for redesign. He highlighted the importance of securing executive sponsorship to align resources and ensure sustained momentum during scaling.

Q: What challenges do organizations face when implementing AI?

Carretero highlighted two major challenges. First, many companies rely on generic AI tools, which are often limited to basic tasks and fail to address nuanced compliance needs. For example, while chatbots can handle routine inquiries, they fall short when tasked with complex workflows like policy optimization or investigative analysis. Organizations need flexible AI solutions capable of customization for advanced use cases.

Second, Carretero emphasized the substantial effort needed to ensure effective change management. AI changes workflows, requiring organizations to train employees to use new tools and adjust how they approach tasks. Generative AI, in particular, demands a mindset evolution, as its capabilities differ significantly from traditional software.

Choi added that organizations often overlook how AI adoption necessitates broader process changes. Transitioning from rules-based to machine learning models, for instance, requires redesigning workflows, recalibrating risk assessments, and retraining investigation teams to interpret model-driven insights.

Q: What are the key compliance areas where AI is gaining traction?

GK asked where the most impactful AI use cases are emerging. Choi highlighted regulatory change management as a high-priority area. With global organizations operating across multiple jurisdictions, AI can simplify the process of aligning regulations with policies and controls. Machine learning models are helping compliance teams identify overlaps, automate control updates, and flag gaps in adherence.

Carretero explained how generative AI improves efficiency in policy and control optimization. It analyzes complex inventories to eliminate duplications and identify controls suitable for automation. He shared examples where organizations reduce compliance costs while maintaining effectiveness by simplifying policy structures.

Q: How can compliance teams measure ROI in AI projects?

GK raised an important question about ROI in AI-driven compliance, acknowledging its challenges. Choi clarified that in compliance, ROI isn’t solely about efficiency; it’s about improving program quality and effectiveness. He cited the example of quality assurance (QA): AI enables 100% case reviews compared to the traditional 10–20% sampling approach. This change enhances program effectiveness, ultimately justifying the investment in AI.

While efficiency gains are important, they should be secondary to the overarching goal of better compliance outcomes. Choi stressed that viewing AI projects completely as cost-saving initiatives risks weakening the depth and quality of compliance programs, especially when facing regulatory scrutiny.

Q: How can AI and human analysts collaborate effectively?

Carretero explained that AI acts as a co-pilot for compliance teams. Automating repetitive tasks like data aggregation and initial screenings enables AI to free up analysts for higher-value activities, such as interpreting findings and making informed decisions.

Choi reinforced this perspective, noting that while AI accelerates processes, it doesn’t replace the need for human judgment. For example, in financial crime investigations, AI can summarize data and detect patterns but skilled analysts are essential for providing context and validating the findings.

Both speakers agreed that this collaboration allows compliance teams to achieve greater efficiency without compromising quality, enabling organizations to confidently address regulatory expectations.

Q: What are the key factors for a successful AI change management strategy?

GK emphasized that AI adoption isn’t just about technology—it’s about change management. He asked Carretero and Choi to share the top three factors for organizations managing AI adoption.

Carretero stressed the need for both top-down and bottom-up engagement. Executive sponsorship is important for advancing AI adoption and process changes, while decentralized experimentation helps identify practical use cases. Identifying internal AI advocates builds enthusiasm and facilitates adoption.

Choi noted that AI transformation takes several years and requires redesigning processes and reskilling teams. Shifting from rules-based to AI-driven systems changes how investigations and compliance reviews operate. 

Organizations must prepare for workflow adjustments to avoid inefficiencies. He also emphasized early regulatory engagement—keeping regulators informed ensures transparency and prevents compliance hurdles. 

With strong leadership, hands-on adoption, and proactive regulatory alignment, AI can drive lasting transformation, not just short-term efficiency gains.

Q: How should compliance teams prepare for future AI adoption?

Toward the session’s conclusion, GK asked the speakers to envision compliance in 2025 and beyond. Choi predicted broader adoption of generative AI and machine learning, with firms moving toward proactive risk management rather than reactive compliance. He emphasized the importance of ongoing pilot programs to continuously refine AI models as technology evolves.

Carretero noted that 2025 could bring notable changes in compliance. AI will move beyond improving current processes to changing how teams approach compliance strategies, helping organizations predict and manage risks.

Wrapping Up

As the session concluded, GK and his guests emphasized the transformative potential of an AI-human partnership in compliance. While AI excels at data-heavy, repetitive tasks, humans provide essential oversight and ethical judgment.

Choi anticipated wider AI adoption as organizations develop use cases and gain more experience with implementation. Carretero noted AI’s ability to move compliance from reactive to proactive risk management, creating a stronger system by 2025.

Lucinity’s Approach

Lucinity reflects the webinar’s focus on combining AI with human expertise, providing a platform that simplifies compliance processes, reduces risks, and improves team performance. Through innovative tools and practical solutions, Lucinity delivers scalable, explainable, and efficient compliance workflows tailored to the needs of financial institutions.

✅ Behavioral Insights Engine - Lucinity analyzes transactional and behavioral data to identify anomalies in real-time, improving financial crime detection while reducing unnecessary escalations.

✅ SAR Automation - Automates Suspicious Activity Report (SAR) creation by collating data, drafting templates, and ensuring regulatory compliance—minimizing the workload for investigators.

✅ Case Management System (CMS) - A centralized, user-friendly platform that consolidates alerts, flagged activities, and case data. Investigators can collaborate, assign cases, and track progress seamlessly.

✅ Feedback Loops for Continuous Learning - AI adapts to organizational needs by integrating investigator feedback, refining machine learning models, and improving false-positive detection.

✅ Dynamic Risk Scoring - Real-time risk assessments prioritize high-risk transactions and customers, enabling proactive monitoring and better resource allocation.

✅ Real-Time Quality Assurance (QA) - AI-driven 100% QA reviews flag inconsistencies, ensuring compliance accuracy and reducing errors across investigations.

✅ Integrated Reporting & Analytics - Customizable dashboards track performance metrics, workload distribution, and compliance trends, empowering teams with data-driven decision-making.

✅ Data Aggregation & Enrichment - Combines structured and unstructured data from multiple sources—transactions, behavioral analytics, and external risk databases—offering investigators a holistic case view.

✅ Automated Case Summarization with Luci - Lucinity’s GenAI-powered copilot, Luci, streamlines investigations by summarizing case details, extracting risk factors, and highlighting anomalies, reducing manual effort.

✅ Adaptive Automation for Risk Monitoring - Machine learning models prioritize alerts and reduce false positives, allowing compliance teams to focus on high-risk investigations instead of repetitive tasks.

✅ Seamless System Integration - Lucinity’s API-first architecture ensures easy integration with existing systems—enabling AI adoption without disrupting infrastructure.

✅ Explainable AI & Transparent Audits - Every AI-driven recommendation includes a clear audit trail, aligning with regulatory expectations and fostering trust in AI-based compliance solutions.

Lucinity integrates automation with human expertise, enabling compliance teams to reduce manual work and focus on strategic analysis and risk management.

Meet The Speakers

David Choi: Partner, Oliver Wyman - David Choi, Partner at Oliver Wyman, has over 20 years of experience in compliance and technology. With a degree in Computer Science from the University of Pennsylvania, his career includes roles at Microsoft and EY, where he focused on regulatory technology and analytics. At Oliver Wyman, he helps organizations modernize compliance operations through advanced analytics and streamlined workflows.

David Carretero: Principal, Oliver Wyman - A Principal at Oliver Wyman, David Carretero has a decade of experience in data science and business strategy. He holds a Ph.D. in Astronomy and Astrophysics from the University of Cambridge and specializes in using data-driven insights and automation to enhance compliance processes.

Guðmundur Kristjánsson (GK): Founder and CEO, Lucinity - Founder and CEO of Lucinity, GK brings over two decades of expertise in financial crime prevention. Before Lucinity, he advanced compliance technologies at Citigroup and NICE Systems. GK’s mission is to provide compliance teams with tools that enhance productivity and simplify decisions when addressing financial crime.

Key Takeaways

The webinar underscored the potential of AI in compliance while emphasizing the importance of thoughtful implementation. Here are the main takeaways:

• Strategic Use Case Identification: Focus on ROI-driven applications that address specific compliance challenges.
• Change Management: Involve both executives and employees to encourage adoption and align processes with AI capabilities.
• Quality Over Efficiency: Prioritize program effectiveness to meet regulatory standards while achieving productivity gains.
• Regulator Engagement: Involve regulators early to ensure compliance and build trust.
• Gradual Scaling: Treat AI adoption as a long-term initiative, starting with pilots and expanding over time.

Adopting these strategies allows compliance teams to leverage AI to simplify workflows, lower costs, and improve decisions. For those interested in exploring how AI and human expertise combine to upgrade compliance, the full webinar is available here.