Fannie Mae and Freddie Mac AI Governance
Understand the GSE AI Rules. Prove How Your AI Is Controlled.
Fannie Mae and Freddie Mac now expect covered mortgage organizations to govern artificial intelligence and machine learning through documented policies, risk controls, oversight, and evidence. The requirements overlap, but they are not the same.
Can you prove what your AI was authorized to do—or what it did?
A policy establishes intent. Execution evidence shows whether a system stayed within approved purpose, data access, tools, actions, and human-approval boundaries.
Direct Answer
What are the GSE AI governance requirements for mortgage lenders and servicers?
Fannie Mae and Freddie Mac both require covered sellers and servicers to govern AI and machine-learning use through documented policies, legal and regulatory compliance, risk management, oversight, review, and disclosure when requested. The effective dates and detailed controls differ, so organizations should evaluate each applicable source separately.
Choose the guidance that applies
Fannie Mae LL-2026-04
Review the Single-Family AI and machine-learning governance expectations for covered Fannie Mae sellers and servicers.
Review Fannie MaeFreddie Mac Section 1302.8
Review Freddie Mac's AI and machine-learning governance, monitoring, audit, accountability, and disclosure requirements.
Review Freddie MacFannie Mae vs. Freddie Mac
See the shared control foundation and the enterprise-specific differences that should remain traceable.
Compare RequirementsCrittora's execution-control layer
From policy to provable execution control
Make authority explicit
Connect an approved use case to the actor, purpose, operation, target system, permitted scope, and applicable time window.
Enforce before commit
Evaluate whether a requested AI-enabled action is within approved authority before it can make a sensitive change.
Produce reviewable evidence
Retain a decision record that helps security, risk, compliance, and technology teams understand why an action was allowed or denied.
Governance must reach the moment of action
An AI inventory identifies where AI is used.
A policy defines the intended guardrails.
A vendor questionnaire records a supplier's representation.
A sensitive action still needs a decision about the actor, operation, target system, purpose, scope, and time window.
Scoped authority should be evaluated before an AI-enabled system calls a sensitive tool, API, or system of record.
Reviewable evidence should explain both allowed and denied actions.
“[A]t a minimum a baseline inventory of its own AI use cases.”
— MISMO President Brian Vieaux, speaking with Scotsman Guide
MISMO 2026 Fall Summit
Let's Meet at MISMO!
Meet Crittora at the MISMO 2026 Fall Summit, August 24–27 in Reston, Virginia.
Bring the AI-governance questions your team is facing. We can discuss how documented policy connects to authority, safeguards, and evidence when AI-enabled mortgage systems act.
Schedule a MISMO ConversationAI Governance Questions
Questions Mortgage Leaders Ask About GSE AI Governance
Each answer stands alone and identifies the enterprise or source it addresses.
Yes. Both Fannie Mae LL-2026-04 and Freddie Mac Guide Section 1302.8 expressly address artificial intelligence and machine learning, so the scope is not limited to generative AI or public chatbot tools.
#gse-rules-cover-machine-learningYes, within the scope of each enterprise's document. Both Fannie Mae and Freddie Mac address covered AI or machine-learning use in mortgage origination and servicing activities connected to their businesses.
#gse-rules-origination-and-servicingA mortgage lender should consider maintaining an AI use-case inventory, accountable owners, documented purpose and manner of use, risk assessments, legal and compliance reviews, vendor records, safeguards, approvals, monitoring results, incidents, and evidence of allowed or denied actions. Exact records should be mapped to the applicable GSE and other legal obligations.
#mortgage-ai-documentationThe official sources should be read carefully before calling a named inventory an express universal requirement. Operationally, a current inventory is a practical foundation for identifying AI use, managing risk, overseeing vendors, and responding accurately when a GSE requests information about AI types, purpose, use, or safeguards.
#gse-ai-inventoryPrimary sources
Use the current official documents when assessing applicability, contractual duties, and implementation decisions.
Crittora provides technology controls and operational information. It does not provide legal advice, certify compliance, or guarantee satisfaction of Fannie Mae, Freddie Mac, MISMO, or other requirements. No endorsement by Fannie Mae, Freddie Mac, or MISMO is implied.
Free Consultation
Schedule a Free AI Readiness Call
Talk with Crittora about your AI-governance priorities, the questions your team is facing, and a practical next step. No lengthy preparation is required.