Abrigo launched Allowance Intelligent Automation on September 16, an add-on to the Allowance product that more than 1,000 financial institutions use for current expected credit loss reporting. It combines automation, AI-generated analysis and workflow controls. The product becomes available by the end of September.

CECL requires institutions to estimate lifetime credit losses on loans at origination rather than waiting for losses to materialize. It replaced the incurred-loss model and has been a sore point for smaller institutions since it phased in.

What It Automates, and Where It Starts

The automation executes configured activities. That covers creating and running calculations, refreshing forecasts, processing pools, calculating qualitative factor scorecards, producing reports and notifying stakeholders when processing finishes. Workflow adds defined tasks, reviews, approvals and reminders.

One boundary is worth marking. Abrigo describes the tool as beginning configured processing as soon as the data is ready. The company is automating the work that surrounds a calculation, not the assembly of the loan data feeding it. That distinction matters, because the data burden is the part community banks complain about hardest.

The release quotes a customer, Lionel Bontemps of Lake Trust Credit Union, saying automation “will let me spend less time on the operational work.” The future tense is accurate. The product has not shipped yet.

The Complaint Is Real

CECL demands more granular, forward-looking data than the model it replaced. Community banks generally lack the data science staff that larger banks assigned to the problem, and they came to the standard last, in 2023, three years after large SEC filers.

The shift to software was already underway. In an Abrigo survey before the deadline, more than half of institutions were still estimating the allowance in spreadsheets, rising to 75% of those under $250 million in assets. Only 16% expected to stay on spreadsheets under CECL.

Even where calculations run in software, executing, reviewing and documenting them stays manual. Month-end and quarter-end compress all of it into a narrow window. The company cites Gartner research putting finance staff productivity among the top three priorities for 88% of CFOs.

That is a genuine pain point rather than an invented one, which is more than can be said for many automation pitches.

Abrigo Has Been Buying This Position

Abrigo serves more than 2,400 institutions across lending, financial crime compliance and analytics. It got there partly by acquisition, adding Valuant and BankLabs in 2022, DiCOM in 2023, TPG Software in March 2024, Integrated Financial Services in 2025 and Journey Technology Solutions this year.

Chief executive Jay Blandford dates the company support for CECL transitions and calculations back more than ten years. TPG is the one that matters here. The Houston investment accounting company was folded into Abrigo’s CECL model, stress testing, income recognition and asset and liability tools. Buying technical accounting capability in 2024 is what makes an automation layer possible in 2026.

The AI Explains, It Does Not Decide

Abrigo launched its agentic platform in July, aimed first at lending and generally available this quarter. The company estimates agentic AI can cut manual labor by more than 40%. The assistants that came before it, in 2025, already spanned lending, allowance, loan review and anti-money laundering. CECL is not a new frontier here.

What distinguishes this release is the role the AI plays. It writes comparison narratives that explain what changed between calculations and flag the drivers. It does not set the allowance. Automation runs pre-configured tasks, workflow forces human review and sign-off, and the model explains rather than concludes.

That is a defensible place to put AI inside a regulated calculation. It is a narrower claim than autonomous fraud decisioning. There, false positives and negatives carry direct cost, and the failure modes are harder to bound.

What to Watch

For a compliance product, adoption matters less than examiner acceptance. Abrigo was named the winner of the 2026 Chartis Award for Managed Services in credit risk. Analyst recognition is not the same as a regulator being comfortable with automated outputs in an exam.

The signal to watch is whether examiners treat these outputs as well-controlled documentation. Banks are reluctant to switch a system once examiners accept it, which is a real moat. It is also the kind that erodes slowly rather than suddenly, as core platforms absorb adjacent functions over successive cycles. Abrigo is not alone in racing that clock. Kinective is assembling a comparable stack for the same institutions from the branch and document side.

Fintechbits covers banking risk technology, regulatory reporting and community financial institutions. Nothing here constitutes financial or investment advice. All analysis represents the editorial views of Fintechbits.