FICO Cuts 15% of Jobs as It Adds AI to Product Development

FICO is reducing its workforce by about 15% while changing how it develops products, giving credit unions a concrete vendor-change event to review across lending, model risk, technology and procurement. The company has not said that the restructuring changes any score, platform, contract or service level today.

In an October 6 Form 8-K, Fair Isaac said management approved the plan October 1. It will reduce organizational layers, simplify the operating structure, optimize processes and tools, and integrate what the filing calls AI-driven product development. Employee notifications began during the week of October 5.

The filing says approximately 15% of positions will be eliminated and estimates about $27 million in pretax fourth-quarter fiscal 2026 charges, primarily for severance and related costs. FICO expects the plan to be substantially complete by the end of its third fiscal quarter of 2027. Reuters reported that the company had 3,811 employees at September 2025, which would put the affected population at roughly 570 if applied to that base.

A vendor change, not a model conclusion

The words “AI-driven product development” do not establish which work will be automated, which products will change or how accuracy, explainability, release quality and customer support will be affected. Credit unions should not infer either improvement or deterioration without product-specific evidence.

The immediate control is to open a material-change review against the institution’s actual FICO dependencies. Inventory the score versions, decision platforms, fraud tools, case-management workflows, APIs, batch files, implementation services and support arrangements in use. Then identify the business owner, contingency path and contractual change-notice rights for each dependency.

That review should request concrete answers: whether product teams or support contacts are changing; whether release, testing or escalation procedures will change; how AI-generated code or analysis is validated; what evidence accompanies material model updates; and whether existing incident, continuity and exit commitments remain intact.

Competition adds a second review lane

The restructuring arrives as mortgage lenders gain a broader credit-score choice. Fannie Mae and Freddie Mac recently opened VantageScore 4.0 to all approved lenders for eligible automated underwriting, while classic FICO remains available. Our implementation analysis explains why a lender must keep the selected score consistent through the credit report, underwriting, pricing and delivery workflow.

That competition does not turn a vendor-management review into an immediate model conversion. A credit union considering another score still needs product-specific validation, fair-lending testing, adverse-action mapping, secondary-market controls and a disciplined transition plan. The appropriate response is to make concentration and model choice explicit, not to substitute one untested dependency for another.

What to monitor next

Credit unions should establish a short watchlist through FICO’s next several release cycles: missed or delayed commitments, turnover among named contacts, support-response time, unresolved defects, documentation quality, model-version notices and any contract changes. Compare observed results with the pre-restructuring baseline rather than relying on corporate statements alone.

The institution’s AI inventory and change-control record can hold the relevant versions, tests and approvals. The vendor-exit playbook provides the corresponding evidence for data return, service continuity and replacement readiness.

FICO has announced an organizational and development-process change. For credit unions, the defensible next step is a measured vendor review that preserves current operations, demands evidence for future changes and keeps model and provider alternatives genuinely executable.