Layla Zulfa
The mortgage industry, a sector perpetually navigating the complex interplay between loan origination and the residual risks that persist even after a loan is sold, is at a pivotal juncture. For AZP Insurance Specialists and its affiliated entities, this challenge is not new; it’s a narrative woven into nearly four decades of mortgage risk and claims experience. The recent surge in Artificial Intelligence (AI) adoption within the industry presents a fresh iteration of this enduring dilemma, demanding a rigorous approach to ensure that technological advancements translate into tangible confidence and financial accountability for lenders.
The landscape of mortgage lending has been shaped by a series of significant economic shifts and regulatory responses over the past two decades. The housing market downturn beginning in 2006, which culminated in the global financial crisis, exposed profound vulnerabilities in underwriting standards, documentation integrity, and risk assessment. This period underscored a critical lesson: the perceived strength of any lending process is directly proportional to the robustness of its supporting evidence and the capacity of all parties involved to honor their commitments when put to the test. The cascading failures, from widespread foreclosures to the near-collapse of major financial institutions, highlighted the devastating consequences of inadequate due diligence and risk management.
In the wake of the crisis, the ensuing decade witnessed a renewed and intensified focus on loan quality, verification processes, and individual accountability. Major players in the secondary mortgage market, such as Fannie Mae and Freddie Mac, undertook significant revisions to their representation-and-warranty frameworks. These adjustments aimed to provide greater clarity regarding lenders’ obligations and the relief mechanisms available in the event of loan defects. For organizations like AZP, this meant grappling with recurring, practical questions: Where did the loan origination process falter? Who bore the responsibility for these failures? And, crucially, how could the resultant financial losses be effectively mitigated and addressed?
The unprecedented refinancing boom of 2020 and 2021, fueled by historically low interest rates, introduced yet another significant challenge: operational capacity. As interest rates began their ascent and the refinancing market contracted, the industry’s focus inevitably shifted towards optimizing costs and enhancing productivity. It is against this backdrop of evolving market dynamics and technological innovation that Artificial Intelligence has emerged as a transformative force. The extensive experience accumulated by AZP across these various cycles renders the questions surrounding AI adoption particularly salient and critical for informed decision-making.
The Evolving Decision-Making Process for AI Adoption
For mortgage lenders, the initial demonstration of an AI-powered technology is merely the starting point of a much broader and more complex evaluation. The fundamental questions revolve around the technology’s seamless integration into existing workflows, the empirical evidence substantiating its performance, and the established protocols for managing exceptions and unforeseen issues. The decisions made during the mortgage manufacturing process, particularly those influenced by new technologies, can carry significant financial implications that extend far beyond the loan closing date.
Key considerations for lenders include the traceability of AI-generated outputs back to their source data, the clear delineation of authority for any modifications made to AI recommendations, and the stringent measures in place to safeguard sensitive borrower information. Furthermore, a critical aspect of this evaluation is defining which responsibilities remain unequivocally with the lender. Technology vendors seeking to deploy their solutions must be prepared to provide credible and comprehensive answers to these inquiries, answers that can withstand the rigorous scrutiny of credit, operations, compliance, and technology departments.
The establishment of a standardized assessment framework is paramount for fostering a common language and facilitating a more efficient adoption process for both lenders and providers. Such a framework enables lenders to pose consistent, pertinent questions, while empowering technology vendors to present evidence that directly addresses clear and well-defined expectations. This structured approach becomes increasingly vital as AI capabilities advance at a pace that often outstrips the traditional procurement and integration cycles of financial institutions.
The timing of AI adoption is critical. It is now influencing decisions that were previously reliant on extensive manual review and human judgment. Lenders, under considerable pressure to improve operational efficiency and reduce costs, require a high degree of confidence that the pursuit of efficiency gains does not compromise visibility into how critical tasks are performed. Similarly, technology vendors seeking widespread market adoption need a demonstrable pathway to prove their readiness beyond the success of a pilot program. A repeatable and robust assessment methodology can effectively guide both parties from initial enthusiasm to a well-informed and confident adoption decision.
The Indispensable Value of AZP’s Extensive Experience
AZP and its predecessor organizations have played a significant role in the mortgage risk landscape, having insured representation-and-warranty obligations on over 5 million mortgage loans, representing an aggregate principal value exceeding $2 trillion. This extensive track record provides a unique vantage point, connecting the evaluation of mortgage technology with the fundamental pillars of underwriting, the exposure to repurchase demands, and the intricate world of claims management.
A mortgage claim, by its very nature, brings the entire origination and servicing process under intense scrutiny. It necessitates a thorough examination of the information available at the time a decision was made, the efficacy of the controls that were intended to govern that decision, and the specific circumstances that ultimately led to a financial loss. This granular, claims-driven perspective is invaluable, informing the critical questions that lenders and technology providers should be asking before embracing new mortgage technologies. It shifts the focus from theoretical benefits to practical, real-world implications and potential liabilities.
Translating Decades of Experience into a Practical Adoption Framework
Recognizing this critical need, AZP has developed the AZP Trust Platform’s AI Mortgage Technology Certification framework. This comprehensive framework is designed to systematically evaluate key aspects of AI governance, data security, internal controls, auditability, operational maturity, and overall risk management practices. It delves into the evidence supporting a technology provider’s readiness for integration into mortgage production environments, with a particular emphasis on how potential exceptions are identified, managed, and resolved.
AZP’s overarching objective is to establish this rigorous discipline as a standard component of the technology adoption process. The aim is to create a shared and objective basis for assessing performance, process integrity, and security protocols prior to the widespread deployment of AI solutions. The certification process is intended to support a lender’s due diligence and internal approval mechanisms, providing a layer of assurance, though it is crucial to note that it does not guarantee specific outcomes nor does it eliminate all inherent risks associated with mortgage lending.
Gateless: A Concrete Example of Certification in Action
Gateless, a prominent provider of AI-driven mortgage solutions, has become the inaugural company to achieve Platinum certification, the highest tier awarded by AZP. Their final assessment, which reviewed evidence up to August 28, 2026, scored an impressive 97.8 out of a possible 100, significantly exceeding the 95-point threshold required for Platinum status. This comprehensive assessment encompassed Gateless’s Smart Underwrite and VeriClear technologies across eleven distinct evaluation categories.
Gateless’s core offering assists lenders in automating significant portions of their underwriting workload, thereby accelerating loan production cycles, reducing manual intervention, and enhancing consistency. This allows underwriting teams to dedicate their expertise to managing exceptions and making critical judgments that require human insight. The company’s innovative approach combines sophisticated agentic AI with established human-defined rules engines and expert systems, applying deep mortgage industry knowledge to the complex tasks of document interpretation and underwriting workflow management. The business objective is to deliver dependable, accurate execution that boosts productivity while preserving essential lender control and oversight.
The exceptional score achieved by Gateless was underpinned by the robust operational discipline embedded within their technological capabilities. Specifically, Gateless earned perfect scores of 99 in Operational Controls and AI Governance, and a near-perfect 98 in Information Security and Performance Analytics. These outstanding results were substantiated by meticulously documented controls, rigorous testing protocols, and tangible production evidence. The assessment identified both areas of exceptional strength and opportunities for ongoing enhancement, reflecting a commitment to continuous improvement.
The calculation audit framework employed by AZP involved a multi-faceted approach, utilizing random sampling, direct review of source documents, and manual recalculation to meticulously test accuracy. This methodology effectively distinguished between actual errors requiring correction and other observational findings, enabling teams to efficiently prioritize issues needing immediate attention from less critical observations. Sampled remediation records demonstrated a clear linkage between identified findings and their routing to product and engineering teams, effectively connecting quality review processes with tangible corrective actions.
This direct connection between risk identification and remedial action is of paramount importance to lenders. Identifying an exception is only the first step; retaining comprehensive evidence, clearly assigning responsibility, and meticulously tracking the response are essential for making the process truly actionable and integrated into daily operations. The AZP assessment specifically recognized this critical interdependency between risk identification, operational review, effective exception handling, and timely remediation.
Furthermore, production reporting provided another crucial layer of evidence supporting Gateless’s certification. Recurring customer reports offered valuable insights into performance trends, adoption rates, automated clearance metrics, and broader operational outcomes. Additional audit trails, quality control data, and performance metrics further strengthened the final assessment. Consequently, Gateless’s score reflects demonstrated, real-world practices rather than relying solely on product descriptions or theoretical control mechanisms.
The review process also incorporated findings from a completed 2025 SOC 2 Type II report. When combined with the mortgage-specific operational testing conducted by AZP, these materials effectively bridge the gap between technology diligence and the critical responsibilities inherent in producing and selling mortgage loans.
The certification announcement highlighted a significant operational achievement: lenders utilizing Gateless’s Smart Underwrite technology have successfully manufactured and sold over 230,000 loans without a single representation-and-warranty claim. While this extensive operating history provides valuable context for the control assessment, it is important to distinguish this performance metric from an absolute guarantee that every loan was defect-free or that future losses are impossible.
For lenders evaluating Gateless, this certification provides a specific and verifiable body of evidence to inform their decision-making. For mortgage technology providers, Gateless’s achievement serves as a compelling illustration of how a meaningful certification can recognize and reward the translation of deep mortgage expertise into documented processes, measurable performance outcomes, and transparently traceable corrective actions.
Uniting Confidence with Financial Accountability
Complementing the independent certification of AI technology is AZP’s specialized mortgage representation-and-warranty insurance program. This program draws upon decades of coverage experience by AZP and its predecessor organizations. Currently, it is provided through a Lloyd’s of London policy, underwritten by Munich Re, and is subject to specific underwriting criteria and policy terms and conditions.
For Gateless, this dual offering—independent certification and available insurance—significantly strengthens its value proposition for lenders. Potential clients can review independently assessed performance metrics and control frameworks, while qualifying clients and transactions can also benefit from financial protection against covered losses. The certification process establishes an evidence-based foundation for trust, while the separately underwritten insurance provides a financial backstop in the event that a covered error leads to a financial loss, subject to the policy’s defined terms, limits, and exclusions.
Ultimately, the successful adoption of new technologies in the mortgage industry must be rooted in trust. This trust, in turn, must be earned through consistent delivery of speed, dependability, accuracy, and demonstrable business results. Gateless’s achievement serves as a powerful example of how proven performance, rigorous independent certification, and accessible insurance solutions can synergistically build and reinforce lender confidence. The opportunity exists for other technology providers to follow a similar path: meticulously substantiating their products, transparently demonstrating their controls, and directly connecting that evidence with robust financial protection for covered exposures. This holistic approach is essential for navigating the evolving landscape of mortgage technology and ensuring long-term stability and accountability.
