The Uniform Appraisal Dataset (UAD) 3.6 update represents a fundamental shift in how residential appraisals are created, delivered and used. Mandated by Fannie Mae and Freddie Mac, the update replaces legacy, form-based reporting with a dynamic, data-driven framework designed for automation, analytics and scalability.
According to Class Valuation Chief Operating Officer Chris Flynn, this is not a simple form revision — it is a structural transformation of the appraisal process. Flynn sat down with Dodd Frank Update to discuss many aspects of UAD beginning with the industry going from “forms to dynamic reports.”
“UAD 3.6 eliminates the traditional library of forms, including the 1004 and 1073, and replaces them with a single, dynamic Uniform Residential Appraisal Report (URAR). Instead of selecting a form, lenders input property characteristics and assignment details, and the report builds itself accordingly,” Flynn told us. “This eliminates incomplete sections and mismatched formats, producing cleaner, more relevant reports for each assignment.”
A Data-First Standard
At its core, UAD 3.6 shifts away from narrative text toward structured, machine-readable data. Each data point is captured in a defined field, allowing it to be searched, validated and analyzed automatically.
Information that once lived in addenda is now embedded in context, improving usability and quality control. Appraisals will now be delivered as ZIP files containing:
- XML data
- A PDF version of the report
- Individual property images
Separating images from the PDF enables direct image analysis and automation, but it also requires lenders to update IT systems that may currently block or limit ZIP files.
Greater Detail in Property Condition
UAD 3.6 introduces more granular condition and quality reporting. Instead of a single rating for the entire property, appraisers now provide:
- Separate interior and exterior ratings
- Room-level detail for kitchens and bathrooms
- Update timelines and condition status
- Structured defect reporting with cost-to-cure estimates
This approach better reflects how properties are maintained and renovated, particularly in high-impact areas such as kitchens and bathrooms.
This added detail increases inspection time. What once took 15 to 45 minutes may now take 30 minutes to an hour and a half, depending on property complexity. While this affects turnaround times and fees, it results in a more valuable and actionable report.
Image Tagging and AI Potential
Photos are no longer appended at the end of a report. Instead, they are embedded within relevant sections and delivered as individual files.
This structure enables AI and computer vision tools to validate condition ratings, identify inconsistencies and detect property deficiencies.
Technology and LOS Readiness
The transition requires significant updates to loan origination systems, or LOS. Beyond supporting a new XML schema, systems must handle ZIP ingestion, new data flows and updated integrations.
The phase-out of ENV delivery adds another layer of complexity, requiring lenders to rethink how appraisal data enters and moves through their systems.
Quality control is shifting from manual review of narratives to automated analysis of structured data. Rules can now flag issues quickly, allowing human reviewers to focus on higher-level judgment.
This change also opens the door to meaningful use of AI in appraisal review.
Operational and Process Changes
Lenders must update engagement letters, which are often tied to legacy form numbers, and reassess internal workflows, quality control rules and underwriting checklists built around those forms.
They should also map the full appraisal lifecycle to identify bottlenecks, particularly around ZIP ingestion, form-based logic and legacy quality control rules.
Historically treated as static PDFs, appraisals are now data assets. Structured data can feed into analytics, risk modeling and automated underwriting, creating new opportunities for insight and efficiency.
Underwriters will need training to adapt to a new way of reviewing appraisals — shifting from reading narratives to validating structured data. While the reports are easier to interpret, they require a different mental model.
Unlike static forms, UAD 3.6 reports are conditional. Sections appear or disappear based on property and assignment characteristics. This requires flexible systems that cannot rely on fixed field positions.
Preparing for the Transition
Flynn emphasizes three priorities for appraisers:
- Start learning now using free GSE resources
- Leverage AMC training and support
- Act early — timelines will not change
“Those who prepare early will transition more smoothly. Those who wait will face increasing pressure as deadlines approach,” Flynn said.