CHICAGO — (January 19, 2026)

Authored by Corey Coscioni, Co-founder Lamina

This is Part 2 of a two-part series on how AI and machine learning are transforming the CLO market. You can ready the first installment here. 

In Part 1, we explored how AI is already delivering pragmatic value across CLO operations. Looking ahead to 2026 and beyond, the conversation naturally shifts to what comes next. Reflecting on OPAL’s CLO Summit in December, one of the most forward-looking aspects of the discussion was how AI can help move the CLO market from a lagging-indicator model to a real-time credit infrastructure. 

Today, CLOs largely rely on delayed data, PDFs, and manual reconciliation. AI creates an opportunity to change that by enabling faster, more accurate, and more connected operational workflows. 

From my perspective, that future includes: 

• Near-instant ingestion of agent updates
• Automated reconciliation across servicers and systems
• Always-accurate portfolio tests and metrics
• Earlier detection of borrower events and amendment ripple effects
• Predictive analytics around funding, ramp and warehouse utilization

Getting there requires moving away from PDFs and manual processes and toward structured data and API-driven connectivity.  In the interim, AI can provide a solution while we wait for the industry to catch-up. 

One point that came up repeatedly—and that I strongly believe in—is that AI is only as good as the expertise behind it. 

Garbage in, garbage out still applies. With AI, it becomes garbage in, garbage amplified. 

Training models without real domain experts leads to scale problems fast. That’s why governance, supervision, and continuous validation are essential as AI adoption grows in CLO operations. 

Data security and regulatory scrutiny remain top concerns for CLO managers and rightly so. 

AI in CLOs should reduce operational risk, not introduce new model risk. From a regulatory standpoint, AI is safest when it is: 

• Deterministic where precision matters 

• Explainable, with clear transformations and validations 
• Supervised, so every output can be reviewed 

At Lamina, we’re seeing customers adopt AI in highly pragmatic, ROI-positive ways. The biggest value isn’t flashy—it’s in closing the operational gaps that slow the CLO ecosystem down. 

Today, our customers use AI to: 

• Extract and structure data from agent bank notices, amendments, and borrower communications 

• Normalize and validate that data across formats and institutions 
• Digitize unstructured communications and integrate them into downstream systems 

The result is less manual work, fewer errors, faster reconciliation, and greater confidence in the data flowing into warehouse desks, analytics platforms, and CLO compliance models. 

We take a bank-first approach to security and compliance, including enterprise-grade security, strict tenancy boundaries, no data commingling, and fully auditable workflows. 

Final Thoughts 

AI isn’t theoretical anymore in the CLO market—it’s already delivering real value. But the firms seeing the most success are the ones treating AI as infrastructure, not experimentation. 

At Lamina, our focus is on helping CLO managers operationalize AI safely—by turning fragmented, unstructured communications into clean, structured, auditable data that can flow across the entire loan ecosystem. If we get this right, AI won’t just improve efficiency. It will help move the CLO industry away from PDFs and delays and toward real-time, data-driven operations that are more accurate, more transparent, and more resilient. 

For more perspectives on AI, data infrastructure, and operational innovation in the CLO market, follow Lamina on LinkedIn.