Home > Articles > AI’s Review of Photos in Appraisal Reports: Big Brother is Watching

AI’s Review of Photos in Appraisal Reports: Big Brother is Watching

By: Francis X. (Rich) Finigan 

20 June 2024

The photographs in your appraisals and the quality and condition rating are about to be scrutinized by AI. Is this just one more step toward hybrid appraisals?  

According to Chris Flynn, head of product and strategy at First American Data & Analytics, “Our AI-powered ACI technology solves two major pain points that AMCs and lenders face – inconsistency between the data and imagery in appraisal reports and the inadvertent inclusion of images containing personal details that should be edited out.” He also expressed, “ACI Sky’s industry-leading appraisal quality control automation paired with the power of Restb.ai’s computer vision capabilities will reduce appraisal revisions, saving time and expense in the appraisal process.”

This new integration leverages AI to extract insights from images captured during the appraisal inspection process to identify inconsistencies in appraisal reports at the time of appraisal delivery, minimizing costly and time-consuming appraisal revisions prior to submission to an appraisal management company or lender.

But, what does this mean? You’ll get alerts about “inconsistencies” in your appraisal report before the AMC or lender sees it, potentially saving you time and money on revisions.

However, could you be correct initially in regards to the property’s condition or quality? Is there a chance that the AI program could be wrong in its assessment? Personally, I trust your judgment over the computer’s!

Even though AI is a great tool, it is just that: a tool. Which is why if you adjust your report based on an ACI Sky notification, I strongly suggest adding an extraordinary assumption. This should state that you’ve relied on ACI Sky’s extensive national evaluations, deeming the information reliable. However, if ACI Sky’s data is found to be inaccurate, it could impact the assignment results.

USPAP SR 2-2(a)(xi) and (b)(xi) require the appraisal to “clearly and conspicuously” state all extraordinary assumptions and hypothetical conditions and communicate that their use “might have affected the assignment results.”

Another inquiry in regards to the development and reliance on this type of technology: Is this just another step toward the use of hybrid appraisals, where the licensed appraiser doesn’t visit the site or take the photos? Instead, an unlicensed inspector does a 7-point inspection, taking photos, which are then sent to an appraiser for a desktop appraisal.

Who are these primary players? 

Restb.ai, a leader in AI and computer vision for real estate, offers image recognition and data enrichment for top industry brands. Its advanced technology analyzes property images, providing insights and property intelligence from 1 million photos daily. More information is available at https://restb.ai/.

ACI, part of First American (NYSE: FAF), provides innovative workflow solutions for real estate appraisers, management companies, and financial institutions. With nearly 40 years of expertise, ACI offers cloud-based software for efficient, compliant operations. More information is available at www.aciweb.com.

First American Financial Corporation (NYSE: FAF) excels in title, settlement, and risk solutions for real estate transactions, leveraging over 135 years of stability, proprietary technologies, and data assets. With $6 billion in 2023 in regards to revenue, it offers diverse services nationwide and globally. Recognized as a top workplace, it was honored by Great Place to Work®, Fortune, and Fast Company. More details at www.firstam.com.

Now that we know who these players are, there is a more crucial question to consider: Who gave them permission to use our data to build these AI models? 

And, on that note…

Good luck and do good work,