The B2B Podcast Index
Index
All categories
MarketingSalesSaaSFinanceHROpsLeadershipCustomer SuccessAI & DataProductStartups & FoundersRevOpsEngineering & DevTools
MethodologySubmit
Best of:MarketingSalesSaaSFinanceHROpsLeadershipCustomer SuccessAI & DataProductStartups & FoundersRevOpsEngineering & DevTools
An independent project byFame
SearchBest episodesGuestsInsightsMethodologySubmit a podcast
Index/Finance/Fintech Hunting
Fintech Hunting artwork

MISMO’s Warning: Mortgage AI Can’t Go Ungoverned

Fintech Hunting · 2026-07-29 · 25 min

0:00--:--

MISMO, the mortgage industry's standards organization, has evolved from a technical back-office concern into a business-critical governance platform under Brian View's nine-month leadership. The conversation centers on two major initiatives: the Mortgage Compliance Data Set (MCD), which standardizes how states request loan data from lenders across all 50 jurisdictions, and Frame - the Framework for Responsible Use of AI in the Mortgage Ecosystem. View emphasizes that most smaller lenders (under 25 in size) lack dedicated AI risk leaders and therefore struggle to govern AI implementation across their organizations, from vendor-provided tools to shadow AI use by employees. The Frame toolkit addresses this gap with three components: a templated policy and procedure document, an AI use case inventory tool, and a risk assessment guide. View stresses that AI governance must be treated as a risk management discipline, not a compliance checkbox, since regulators will hold lenders accountable regardless of whether humans or AI agents make lending decisions. MISMO's upcoming August summit will feature hands-on Frame workshops, and the organization is rolling out two new certifications: one for governance advisory firms and one for AI vendors, helping lenders evaluate which AI products meet governance standards.

Key takeaways

  • →Smaller lenders (under 25 in size) typically lack dedicated AI risk leaders, leaving them vulnerable to ungovened AI implementations from vendors and internal shadow AI use.
  • →MISMO's Frame toolkit provides a three-part structure for AI governance: policy templates, use case inventory, and risk assessment tools that lenders can deploy in 72 hours if regulators request proof of oversight.
  • →AI governance is a risk management discipline, not a compliance exercise - existing lending regulations like fair lending and redlining apply whether humans or AI makes the decision.
  • →The MISMO business glossary of 8,000+ standardized terms enables consistent data inputs for AI models, directly improving output accuracy and interoperability across lender tech stacks and investors.
  • →MISMO is launching AI Governance Advisory and AI Vendor certifications to help lenders evaluate and implement AI solutions with confidence that vendors meet industry governance standards.

Guests

Brian View

Topics in this episode

MISMO (Mortgage Industry Standards Maintenance Organization)Mortgage Compliance Data Set (MCD)AI governance certification for vendorsAI Governance Advisory certificationMISMO business glossary (8,000+ defined terms)MISMO reference modelResidential Board of Governors (MBA ResBOG)AI Community of Practice (AICOP)Fannie Mae and Freddie Mac AI guidance

Questions this episode answers

What is MISMO's Frame toolkit and what does it help lenders do?

Frame is a Framework for Responsible Use of AI in the Mortgage Ecosystem consisting of three tools: a templated policy and procedure document, an AI use case inventory tool to identify where AI is being used (including shadow AI by employees and through vendors), and a risk assessment tool to evaluate each use case. It allows lenders to document their AI governance in 72 hours if regulators request proof of oversight.

Why do small and mid-size lenders struggle with AI governance compared to top-10 lenders?

Smaller lenders typically lack a dedicated chief risk officer or AI risk leader, with risk responsibilities often rolling up to a COO or CEO instead. This means they have limited organizational capacity to evaluate vendors, track AI implementations, or manage compliance risks compared to larger lenders with dedicated risk armies.

Is AI governance the same as AI compliance?

No. AI governance is a broader risk management discipline that falls under general risk governance, while compliance is a specific subset of that risk. Existing lending regulations haven't changed - fair lending and redlining rules apply whether humans or AI make the decision - so lenders shouldn't wait for new AI regulations before implementing governance.

What is the Mortgage Compliance Data Set (MCD) and why does it matter?

MCD standardizes how the 50 states request loan data from lenders for audit purposes, replacing the previous system where each state had its own checklist and format requirements. This reduces burden on multi-state lenders and enables technology vendors to build solutions for multiple states rather than custom integrations for each.

How do MISMO standards help with AI model accuracy?

Using MISMO's business glossary of 8,000+ standardized terms as a data foundation ensures that inputs fed into AI models are consistent and accurately defined, which directly improves the consistency and accuracy of AI outputs across the organization.

Conversation analysis

Computed from the transcript - who did the talking, and the words that came up most.

Share of words spoken

  • Speaker B77%
  • Speaker A23%

Most-used words

mismo35industry22data22lenders21risk14governance14compliance10frame10start9lender9back8model8question8mortgage7standards7point7

Episode notes

Mortgage AI is already inside lender workflows, vendor platforms and employee tools. But if a regulator, investor or business partner asked where AI is being used, who owns the risk and how those decisions are governed - could your organization provide a clear answer? In this episode of the FinTech Hunting Podcast, Michael Hammond sits down with Brian Vieaux, President of MISMO, to examine the growing gap between AI adoption and responsible AI governance in the mortgage industry.

Full transcript

25 min

Transcribed and scored by The B2B Podcast Index.

Speaker A: Ladies and gentlemen, welcome to a new episode of the FinTech Hunting podcast. We've got an awesome guest for you today. He's a mortgage industry veteran, he's an author, he's a standards leader, he's one of the industry's most influential voices on AI data, standards innovation and the future of homeownership. Please help me welcome back to the show Brian View, president of mismo. Brian, it's great to have you back.

Speaker B: Yeah man, thanks for uh, having me. I feel like uh, you and I have done this a few times, so I'm looking for my five year jacket like they do on Saturday Night Live.

Speaker A: I gotta start working that out. You have been on the show, you always provide a lot of great insights. Tack even on the side we talk about what we're using for the podcast and everything like that. But for our listeners, okay, we're right at the end of July. Give everyone kind of a state of the union of. We're where Mismo is at in kind of the year, year and a half that you've had now at the leadership helm.

Speaker B: Not even a year. You're aging me fast, but uh, I'm at like nine months. Um, yeah, it's been um, there's a lot that's gotten done and um, you know, a lot of ongoing work. So at any given time there's 30 or so active work groups. And these work groups consist of industry experts, uh, that volunteer their time, you know, expertise, experience, etc. To solve some problem for industry. And that's really the basis of mismo. We are for the industry, by the industry. Uh, anybody in industry can submit a work idea or a problem statement and then as you know, we put it through a vetting process and if it gets stood up as an actual work product then we do a call to participate and anybody from industry can come in and participate in collaborating on a solution. So there's 30 or those right now that are in some status if you will. The two, uh, probably most uh, visible to industry is one of them is the Mortgage Compliance Data Set, also known as mcd. This is standardizing uh, how states request uh, loan data from lenders or licensees for loan selection, ultimately for audit. So pre McD, think of 50 states. Each has its own checklist, column headers, row labels, blah blah, blah, blah blah. Lenders that are multi state are just getting inundated for essentially the same data but they have to put it together in all these different formats. So MCD attempts to kind of create a standard and bring that together uh, helps lenders, improves, uh, quality, also helps, um, the state regulators, um, when you think about standardizing this data set now, technology can come in and create, uh, solutions that they probably wouldn't do for a single state, but they might do it for multiple states.

Speaker A: Right. Well, talk to me also. You've done a fantastic job in my opinion. Maybe that's why I thought because you've accomplished so much that it was a year, year and a half. But I think one of the things that's jumped out at me is the fact that you've brought Mismo more mainstream for the business user, not just the, you know, back in the day, I look back 15, 20 years, I won't go any further and date myself. But when I look at that, Mismo was kind of for the super techies, uh, the people that were really in the nitty gritty and doing, you know, the code and everything like that. But you've made Mismo much more accessible and really brought a great ROI to the common person in the industry. Talk to me a little bit about that.

Speaker B: Yeah, well, first of all, it's, it's folks like yourself that are, you know, gracious to invite me on a platform like yours to share, uh, you know, what we're doing at Mismo. And that's, you know, when I started with, uh, the organization, call it nine months ago, I set off, off on a campaign of three A's, awareness, adoption and alignment. And this is the awareness part, just getting out as frequent and as often and as many different places as possible to share what's happening at Mismo so that to your point, we can start to reach different segments and cohorts of industry, um, that aren't or weren't necessarily engaged with mismo. Going back to McD, it's a great example of that particular work product brought in compliance and risk people that hadn't prior been engaged with Mismo. So kind of expanding, you know, in different parts of the business, um, and really been focused on helping lenders of all shapes and sizes, even the smallest of lenders, really understand the role Mismo plays in driving efficiency and ultimately helping us all kind of tackle this cost to produce issue that still is a pretty massive one for the industry. Um, and then of course, um, with how fast AI is moving, I don't have to tell you about it. You're right in the middle of this. Um, you know, we're seeing, uh, a lot of new players come into the mortgage vertical with point solutions and other tools that are, you know, AI powered or AI Native and Mismo, um has been a place where a lot of those products and companies have sought or come into, you know, being, I guess being smart enough to know if they're going to build something, they should probably connect to this language that's spoken across the industry. And we're seeing a lot of that.

Speaker A: Well, and I think they have faster adoption without necessarily being in the industry for 15, 20, 25 years. So I think that helps. The other thing I'm observing as we talk about Mismo and the NBA, I see a much greater willingness between the NBA and all of the other industry participants. You see a, ah, great collaboration with the NBA and HousingWire or NBA and ICE or whoever it may be. And I think that really helps, giving you guys a platform to tell your story and really show the benefit.

Speaker B: Yeah, it's. First of all, NBA is, is an amazing partner of Mismo. Full, full transparency. We are a wholly owned subsidiary of mba, um, so you better hope they're a great partner. But notwithstanding that relationship, they are amazing. Uh, anytime that, um, the membership team or the policy team M or uh, the technology team led by Rick Hill can get Mismo in front of members, they are doing it. So they're a great advocate for the work we do at Mismo, um, Housing Wire. Always longtime fan of Clayton and the team there. Clayton's actually moderating a panel at our upcoming Mismo summit. It's our super. Our credit super panel where we're going to have executives From Experian, Equifax, TransUnion, Vantage, Score and FICO all on the stage together, and Clayton's going to moderate that conversation.

Speaker A: Um, is that a cage match or.

Speaker B: Pardon me, is that.

Speaker A: Is that a cage match?

Speaker B: No, no, no, no, that's. No, no, not at all. This is, this is the beauty of Misma. We are, we are, uh, a place where collaboration is encouraged and frankly, where it happens. And we already see it today in our credit reporting work group where those companies I just mentioned work together to solve problems for industry. So we'll leave the.

Speaker A: Glad you highlighted that because that is important, because everyone hears and there's so much talk and battling back and forth. But I think to your point, if we're ever going to have problems, progress, they've got to come. And if you guys can deliver a, uh, collaborative environment, kudos to you.

Speaker B: What's. What's cool, Michael, is, is, you know, I've seen this now in this. In the nine months in this seed. And you know, I was like a lot of people in our Industry that, you know, I wasn't like you, involved in Mismo, you know, ten plus years ago. Um, I've known about it and you know, if you really quiz me, I probably could tell you what all the letters stood for, but maybe I needed to cheat a little bit. Um, but now being inside and actually seeing how the sausage is made, um, this is truly a place, and I'll say place because it's physical and digital, right? This is a place where competitors truly take off their company logo and put the industry logo on and you know, really work closely together, uh, leveraging, you know, the Mismo reference model and the tools here to create solutions that benefit everybody. Uh, and it's really, uh, it's fun to watch it happen. It's, uh, it's, it's encouraging. Uh, and we'll leave, you know, we'll leave other venues for those cage matches because, you know, they're fun too.

Speaker A: You briefly touched upon AI. AI is already making or influencing decisions across the mortgage life cycle. We know that. We're not hiding from that. Are most lenders truly governing it or are they trusting vendors and hoping nothing goes wrong?

Speaker B: Wow, great question. And that thing's loaded for bear. I, um, would tell you, um, I've seen In the last 90 days a real, uh, change in the mindset of lenders. You know, we could look at the top five, 10, 15 lenders. They all, they have robust, mature risk governance. You know, a lot of them have like an army of people that are managing risk and compliance. So they've, they're doing their thing and they're going to, they're going to get it right. Uh, no question about it. As soon as you start to dip below, you know, into the 25 and below, and certainly the mid to small size lenders look at a lot of those companies don't have a chief risk officer. Right. A lot of times the role of risk rolls up to a coo, maybe even a CEO or president. Um, and then specifically as it relates to AI, um, those lenders m. If they don't have a risk leader, they definitely don't have an AI risk leader. And so this is an example of where MBA and Mismo, uh, really linked up. So the resbog, which is the Residential Board of Governors of mba, they set kind of the priority for each fiscal year for mba. Last year at MBA annual, the Resbog, led by Dan Sugg, a Michigan, uh, friend of ours.

Speaker A: That's right.

Speaker B: Um, Resbog prioritized AI governance as its top two or three things that it wanted to deliver for members. And it came to MISMO to um, kind of do the work where we then took an existing community of practice, our AI community of practice, and created a subgroup within that AICOP to focus on building, uh, uh, a toolkit for lenders. And I'm proud to say, about a month ago, a little over a month ago, we published V1 of frame. So think of Frame as an acronym, the Framework for Responsible Use of AI in the Mortgage Ecosystem.

Speaker A: Yep.

Speaker B: And what Frame is? This is a long winded answer, but hopefully this will help. What Frame is, is a set of tools. Version 1 is a templated policy and procedure document. It's very robust, probably more robust than most small lenders would need. So we encourage lenders that download the toolkit to edit that particular document to meet their company needs. Uh, but it is very robust. And then the second tool is an AI use case inventory tool. And so this is where I think most lenders miss, is that they don't know or they haven't identified and they certainly haven't documented all of the use cases where AI is impacting their business either directly or, or indirectly through vendors more often. And then there's this whole shadow, uh, AI where employees like you and I are using AI that may or may not be enterprise grade, right?

Speaker A: That's right.

Speaker B: So that inventory is a place to gather all those use cases. And then the third tool is a risk assessment tool where the lender is literally guided to perform a risk assessment on each of those use cases. So now if you think about this toolkit, if you implement, if you develop that policy and procedure, do your inventory and then start to create your, uh, risk assessments on your use cases, when someone comes knocking, whether it's Fannie Freddie, an investor that buys your loans, or the state and starts asking you what you're doing with AI and your oversight of it, you now have something you can produce like, like that, Right. Uh, James Brody talks a lot about you don't have 72 days to produce this. You have 72 hours and you better be doing the work point well.

Speaker A: And that was going to lead right into my next question of, with mismo, Frame, it's designed to bring structure to responsible AI adoption. Right?

Speaker B: Yeah.

Speaker A: Uh, so with that, what's the biggest mistake companies are making when they treat AI governance as a compliance exercise instead of a business discipline?

Speaker B: Yeah, I would say, um, you need to have your eye on compliance. And what we talk about a lot, and you'll hear Rick Hill, especially from mba, talk about AI governance and frame in particular is really a risk governance. Uh, it falls under risk governance at a very high level. Compliance, certainly compliance risk is a real thing as you think about AI, but nothing's changed from a lender's perspective in terms of, you know, the regs that they're responsible for and that they're upheld to. AI is another set of tooling that certainly could impact it. And if you get something wrong and it's, it could get wrong fast at scale.

Speaker A: Right.

Speaker B: Um, but, but compliance hasn't really changed. You're still as a lender responsible for all the compliance stuff. You were. And that's why we kind of want to take it a level up and talk about, know, AI governance as part of a risk governance, uh, discipline.

Speaker A: That's such a good point. And I was on a panel with Rick and in the audience somebody kind of made the point of, well, we're waiting for the regulators. And he's like, we don't need more AI regulation. We need the governance and the policies and the procedures. Fair lending is fair lending. Whether Brian does it or Brian's AI agent does it.

Speaker B: Yes.

Speaker A: You're, uh, still on the hook for doing that properly. Redlining, red light. So he made a really good point. I'm glad you brought that up. Of, uh, don't wait for the regulations. And then you're going to finally start doing something. Start incorporating the policies, procedures, the governance so that you're adopting it properly. But lending is lending. We have enough regulations out there. We don't need a ton more.

Speaker B: Exactly. And if you. I've talked to a ton of lenders and I guess if you did one thing and you had to pick one thing to do as it relates to AI governance, I would say start with the inventory. Literally inventory all of the use cases. Because at a minimum, if you can at least answer the question where you're using AI, you're not going to get the side eye from the regulator.

Speaker A: Right. Well, and like you said, having that library at your disposal to say I got to start, I never even thought about this vendor could have been using it, or this person, or internally. All right, we've had a lot of great discussion. Now I'll ask you a tougher, more challenging, uncomfortable question. But I know I've known you for a long time and I know you'll nail this. So the uncomfortable question is for mismo, if standards are so valuable, why has adoption been so historically slow? And what as an industry, do we need to either stop tolerating or doing more of with these standards.

Speaker B: Such a, such a great question. And it's the question that I, you know, was asking myself as I was vetting the opportunity and I was being vetted as a candidate. Um, and when I. The best way to answer it is, um, let's just say mid to small size lenders, that that's the bulk of the mortgage companies in our industry, most of them, if as they think about MISMO and data specs and data standards, they're relying on their tech stack, really. Right. Um, until today as it relates to mismo, they're still relying on their tech stack. But now today every lender of any size should have a data strategy. Correct in. The data strategy can't just be my single source of truth is my los, because that's true as it relates to the loans that are going through the manufacturing process. But what about loans, uh, that never make it in or historical loan data and all the other stuff that's around your business. And so I think with one of the places where AI is really helping level the playing field is there's enough tooling now, it's making it more accessible for smaller lenders to have a true data strategy. Um, and so if you're going to have a true data strategy and build a data model of any size and scale in your organization, you need to start with something, uh, a set of definitions, let's say.

Speaker A: Right.

Speaker B: And so Mismo, uh, has this business glossary of over 8,000 defined terms for our industry that's been around for two decades plus. It's, we have a dedicated work group that works on this every month. They're looking at potential new definitions. Some go away, some are tweaked and change, etc. And so if you're building a data model, I think this is where MISMO is important today because we're all interconnected and we have this interoperability, right. Lender to tech and lender to lender. If we're selling loans, if you're building a data model, you want to ultimately make sure that your data can talk to your tech stack, your investors, your compliance and your regulators. Even so at its core, if you use MISMO as your data foundation, it's going to accelerate your ability, um, to build something that you can actually connect with others on. The second, uh, thing that I think makes most sense for mismo, with every lender and it goes right, uh, with this data strategy, if you're going to leverage AI, which everybody is, it's just a reality, one form or another.

Speaker A: Absolutely.

Speaker B: Like any Model, you want to make sure that the inputs are consistent and accurate. And so this is where standards matter. And so if you're leveraging the MISMO reference model, the business glossary, and now you're starting to apply AI in different parts of your business where you're connecting your data, knowing that you've got consistent defined data elements is going to allow you to get more consistent, accurate output from your AI. So I think for a lot of years, just to kind of reiterate, lenders, smaller lenders relied on their, their tech providers and now I think with, with you know, standing up a data model, managing a data model and certainly leveraging AI, you know, leveraging standards and, and defi, you know, a business glossary of 8,000 terms makes a lot of sense.

Speaker A: Couldn't agree more. And I think you did a fantastic job of explaining that. Brian, you and I could talk forever. What's next for Mismo?

Speaker B: So, um, we are um, we hold, we host three events every year. So we're three weeks from hosting our August summit, August 24 through 27 in Reston, Virginia. A lot of focus on AI at that summit. Uh, on Monday the 24th, uh, we're doing a four hour workshop where we're going to really go deep into Frame and teach workshop uh, attendees how to use the tools. So we're excited about that. Uh, we're going to announce. I'm going to do it here on your, your show. By the way, um, I've always loved uh, the fintech hunting brand. I think it's like one of the smartest podcast names. It's, it's amazing.

Speaker A: I got to give credit to Eric Quiela. He, him and I were sitting in at lunch when we came up with that title.

Speaker B: You know what they say, a blind squirrel every once in a while will find a nut. So good, good job Eric. Um, the, the things that we're uh, going to announce at Summit, uh, is around certifications. So Mismo has a bunch of certifications. We have certified consultants, certified products, etc. Etc. A lot of our certs in the product side are in the evault E Mortgage kind of space.

Speaker A: Okay.

Speaker B: We're going to stand up two new certifications around AI. One is an uh, AI Governance Advisory certification. So this is where m, you know, firms um, that are like Brody Gap as an example, um, who are active in that, in that governance space can be certified against Frame. So we're going to be certifying these advisory firms that they're you know, effectively uh, certified implementers of the Frame toolkit.

Speaker A: Love it.

Speaker B: So that's number one. We're excited about that. We have about a dozen companies that have um, uh, already kind of pre applied if you will. And then the second certification is going to be at the vendor level where we're going to do uh, an AI governance certification for vendors. So when we talk about, you know, a lot of the hesitancy on lenders parts around AI, it's not knowing how to evaluate their vendors to make sure that the vendor's use of AI is not going to get the lender in trouble.

Speaker A: Right.

Speaker B: Uh, because I think you know this, both Fannie and Freddie, their, their guidance, their guidelines put the onus 100% on the seller servicer.

Speaker A: Exactly.

Speaker B: And so we, we believe by uh, mismo certifying vendors it'll help lenders um, get more comfortable um, exploring different AI products and use cases and companies knowing that, that you know, at a minimum they've gone through this certification process. So those are two things that we're really excited to bring to the industry because we think it's going to add tremendous value.

Speaker A: Well, you are a wealth of knowledge and insights. You need to come back maybe quarterly we can get you back and those MISMO updates for those of you, if they want more information, if they want to dive deeper, if they want to see some of these certifications, where is the best place that they can go?

Speaker B: So you can obviously hit me up on LinkedIn. That's kind of where I live and breathe most days. Um, all my contact information is on my LinkedIn profile. You can go to mismo.org and you can navigate to uh, get involved and there's a dropdown where you can go to any one of our 30 work groups. You can hit that. We have a landing page for frame, you know the framework, uh, there. Um, and of course we, you can link to our summits and check those out as well.

Speaker A: Well, I can't thank you enough for being a guest on this episode of the fintech hunting podcast.

Speaker B: Thank you sir.

More from Fintech Hunting

All episodes →
  • Waiting for Rates to Save You? KP Says Mortgage’s AI Race Has Already Started89 / 100
  • Mortgage Lenders Are Missing the Inventory Engine | Dana Georgiou on Private Lending61 / 100
  • AI Won’t Replace Relationships - But It Will Expose Who’s Been Faking Them | Fintech Hunting Podcast51 / 100
  • Mortgage Has an AI Problem Nobody Wants to Admit | Trish Maraski48 / 100
  • Michael Hammond, Founder of NexLevel Advisors, is the leading fractional CMO in mortgage and mortgage technology, specializing in AI-powered growth strategy and audience development.
Explore the best B2B Finance podcasts →
All Fintech Hunting episodes →