As we move through 2025, non-bank lenders face an increasingly complex regulatory environment characterised by fragmentation and shifting enforcement priorities. While federal oversight through the Consumer Financial Protection Bureau (CFPB) has diminished, state regulators have stepped forward to fill the void, creating a patchwork of compliance requirements that vary significantly across jurisdictions.
In Michigan, Attorney General Dana Nessel has reaffirmed her office's commitment to enforcing consumer protection laws, highlighting that the CFPB has "won back for American consumers more than $20 billion, as direct payments back to wronged customers or into relief funds for impacted victims." Nessel emphasised that her office's Consumer Protection Team would "continue to enforce the Michigan Consumer Protection Act to hold businesses accountable and defend consumers from illegal and abusive business practices" [1].
Similarly, New York regulators have intensified their consumer protection efforts. Adrienne A. Harris, the state's Superintendent of the Department of Financial Services, announced an expansion of the state's regulatory capabilities with a simple but telling statement: "We're hiring" [2].
Despite these shifts at the federal level, foundational statutes like the Equal Credit Opportunity Act and Dodd-Frank Section 1071 remain in effect. The CFPB has even issued recommendations to states for enhancing their consumer protection frameworks, suggesting the incorporation of "abusive" practices into state laws, strengthening investigative tools, and creating clear prohibitions on junk fees [3].
For lenders operating across multiple jurisdictions, this regulatory fragmentation creates significant challenges. Each jurisdiction may have different enforcement priorities, reporting requirements, and compliance thresholds. The resulting complexity demands sophisticated tools that can adapt to varying requirements whilst maintaining operational efficiency.
The financial services industry is experiencing rising delinquencies, with 40% of financial institutions reporting an increase in delinquencies over the past 12 months [20]. This trend is prompting a reassessment of traditional credit assessment methods and driving the need for more sophisticated compliance frameworks that can handle increased risk whilst maintaining regulatory adherence.
Fast-growing fintech companies are increasingly adopting an AI-First-Approach to enhance compliance processes, analyse large volumes of regulatory data, and maintain profitability whilst meeting stringent regulatory requirements. This approach represents a fundamental shift in how lenders approach compliance—from a cost centre to a strategic advantage.
The transition towards analysing large volumes of data is driven by a fundamental difference in business models. Traditional banks typically focus on large-scale loans, whilst alternative lenders often concentrate on smaller, more frequent transactions. This difference necessitates different approaches to compliance and risk management [4].
Machine learning methods can be used to estimate each client's lifetime value and evaluate the potential for long-term cooperation. This allows companies to balance desirable conditions for customers, attract new users, and maintain strong profitability—all whilst ensuring regulatory compliance [4].
What makes this approach particularly powerful is its ability to adapt to changing regulatory requirements. As regulations evolve—which they inevitably will—AI systems can be trained on new datasets, identifying patterns and relationships that might not be immediately apparent to human compliance officers. This adaptability is crucial in a regulatory environment where requirements can vary significantly across jurisdictions and change rapidly.
"The integration of digital platforms, machine learning, and large language models in financial services has enhanced accessibility but also introduced regulatory challenges. These include concerns about data privacy, algorithmic bias, financial stability, and consumer protection, necessitating evolving regulatory frameworks." [10]
The regulatory challenges associated with ML and digital platforms in financial services extend beyond basic compliance concerns. As financial institutions increasingly rely on these technologies, they must navigate complex issues related to algorithmic transparency, data governance, and model risk management. Regulators are particularly concerned about potential biases in ML algorithms that could lead to discriminatory lending practices, as well as the systemic risks that might arise from widespread adoption of similar algorithmic approaches across the industry [10].
The adoption of AI in financial operations is accelerating rapidly. According to Gartner, 58% of finance functions are utilising AI in 2024, marking a 21 percentage point increase from 2023 [21]. This surge demonstrates the growing recognition of AI's potential to transform compliance operations from reactive cost centres to proactive strategic advantages.
By treating compliance as a strategic advantage rather than a burdensome cost, forward-thinking lenders are positioning themselves for sustainable growth in an increasingly regulated industry. The question is no longer whether to invest in AI-powered compliance tools, but how to implement them most effectively.
ML algorithms are transforming compliance by automating routine tasks, reducing false positives in fraud detection, and providing predictive insights into potential compliance issues before they become regulatory violations. This shift allows compliance teams to focus on strategic initiatives rather than manual reviews, turning what was traditionally seen as a cost centre into a source of competitive advantage.
AI-driven risk analysis has become the new standard in financial strategy, with algorithms integrating unstructured data into adaptive models that evolve in real time. This approach has significantly increased fraud detection rates while reducing false positives. These systems evolve through human insights to enhance accuracy over time, with AI capable of reducing compliance costs by up to 40% and improving investigation throughput [22].
The shift from reactive to predictive compliance is where the real transformation occurs. As highlighted in recent industry analysis, automated payment investigations help resolve errors faster and prevent them from happening again. AI-powered automation built into alert investigation platforms streamlines case management, shortens resolution times, and reduces the need for rework, while maintaining audit-readiness and compliance integrity [23].
Specific applications of ML in compliance that have delivered strategic advantages include:
Machine learning algorithms are strengthening fraud prevention and streamlining identity verification processes, helping lenders meet Know Your Customer (KYC) and Anti-Money Laundering (AML) requirements whilst reducing false positives and improving customer experience.
The financial sector is facing substantial losses due to sophisticated fraud schemes. In 2023, banks globally lost $230 billion to fraud schemes and scams, whilst financial institutions spent $485 billion on financial crime compliance and fraud prevention in 2024 [24]. These staggering figures underscore the critical need for advanced ML-powered fraud detection systems.
ML-driven limit strategies like Recommendation Systems have gained significant traction across the lending industry. These systems analyse patterns in transaction data, customer behaviour, and external information sources to identify potential fraud whilst minimising false positives that create friction in the customer journey [4].
The integration of ML into fraud prevention and identity verification processes represents a significant advancement in regulatory compliance technology. Traditional rule-based systems often generate excessive false positives, creating unnecessary friction in the customer journey and increasing operational costs as compliance teams manually review flagged transactions.
Know Your Customer (KYC) is a sophisticated compliance framework used by financial institutions to identify and validate client identities, transactional behaviors, and risk exposures. KYC is designed to mitigate systemic vulnerabilities such as identity fraud, illicit financial flows, and potential terrorist financing. It involves a structured process of gathering detailed identity evidence and applying complex risk-scoring models to ensure customer authenticity, credibility, and behavioral consistency [25].
Advanced KYC paradigms now leverage AI/ML for document verification with OCR (Optical Character Recognition) and NLP (Natural Language Processing) for facial recognition, enabling institutions to move from point-in-time KYC to perpetual KYC (pKYC). This continuous monitoring approach allows for real-time anomaly detection, significantly enhancing compliance effectiveness while improving customer experience [25].
A particularly promising development is the integration of blockchain technology into Loan Origination Systems (LOS). Blockchain establishes trusted records of transactions, documentation, and events, enhancing transparency, security, and automation across the lending lifecycle. This integration is proving especially valuable in reducing risk and combating fraud [11].
ML algorithms, by contrast, can analyse patterns across vast datasets to identify potential fraud with greater accuracy. They can distinguish between genuine anomalies and normal variations in customer behaviour, reducing false positives whilst maintaining or even improving fraud detection rates.
This capability is particularly valuable in the context of KYC and AML compliance, where the consequences of non-compliance can be severe. By automating identity verification and transaction monitoring, ML-powered tools help lenders meet regulatory requirements more efficiently and effectively.
Financial institutions are increasingly adopting overlay-based anti-money laundering (AML) models, which enhance existing systems rather than replace them. These models add intelligence layers to improve performance while maintaining operational continuity. Modern AML overlays powered by AI and machine learning enable faster detection, improved prioritisation, and clearer visibility into risk. A standout innovation in this space is federated learning, allowing AML systems to evolve collaboratively across institutions without compromising data privacy [17].
As lenders increasingly rely on ML algorithms that require vast amounts of data to function effectively, they face growing challenges in balancing innovation with data protection compliance. This tension requires sophisticated compliance tools that can enforce privacy-by-design principles whilst still enabling the data analysis necessary for effective ML implementation.
ML tools must be carefully designed to comply with relevant data protection laws, including the General Data Protection Regulation (GDPR) [4]. This presents a significant challenge, as ML algorithms typically improve with access to more data, whilst data protection regulations aim to limit data collection and processing to what is strictly necessary.
To address this challenge, lenders are implementing privacy-enhancing technologies (PETs) that allow for data analysis without compromising privacy. These include techniques such as:
In practical application, federated learning has emerged as a standout innovation in the AML space, allowing systems to evolve collaboratively across institutions without compromising data privacy [17]. This approach enables financial institutions to benefit from collective intelligence in fraud detection while maintaining strict data protection compliance.
For example, overlay-based AML models that integrate with existing systems via APIs or batch interfaces allow institutions to see a return on investment within months rather than years. These systems enable earlier identification of high-risk activity, improved signal-to-noise ratios in alerts, and reduced workload for compliance teams through smarter triage and scoring—all while maintaining data protection compliance [17].
Differential privacy techniques are being implemented in credit scoring algorithms, allowing lenders to gain insights from large datasets while ensuring that individual borrower information remains protected. By adding carefully calibrated statistical noise to the data, these systems prevent the identification of specific individuals while maintaining the accuracy of aggregate analysis.
Recent research has introduced a framework combining quantum deep learning techniques with adaptive modeling for Row-Type Dependent Predictive Analysis (RTDPA) in credit risk assessment. This hybrid approach aims to enhance both the accuracy and efficiency of evaluating credit risk in the banking sector while maintaining data protection compliance [12].
By implementing these technologies, lenders can leverage the power of ML for compliance purposes whilst ensuring that they themselves comply with data protection regulations. This balanced approach is essential for sustainable innovation in the lending industry.
The most effective privacy-enhancing technologies for lending compliance include federated learning for distributed data analysis, differential privacy for protecting individual data points whilst maintaining statistical validity, and synthetic data generation for testing compliance systems without exposing real customer information. The optimal approach often combines multiple technologies tailored to specific regulatory requirements and data sensitivity levels.
Advanced compliance technologies are becoming key drivers of competitive advantage, with institutions that effectively harness regulatory and transactional data building trust, acting with agility, and scaling confidently in a rapidly changing financial ecosystem.
Technologies like rt360, which monitors over USD 350 billion in assets, have resulted in savings of at least USD 500 million per bank annually [5]. These substantial cost savings demonstrate that ML-powered compliance tools deliver measurable ROI beyond just risk mitigation. rt360's real-time monitoring capabilities and predictive analytics have been particularly effective in helping banks identify potential defaults earlier and take proactive measures to mitigate risks. The system's ability to analyse vast amounts of transactional and customer data in real-time enables financial institutions to detect compliance issues before they escalate into regulatory violations [5].
The implementation of rt360 has been particularly effective in responding to India's non-performing asset (NPA) levels, which reached 7.48% in FY16 and remained high at 7.47% in subsequent years. By providing real-time monitoring capabilities and predictive analytics, the system has helped banks identify potential defaults earlier and take proactive measures to mitigate risks [5].
The ROI of ML-powered compliance tools can be quantified in several ways:
Beyond cost savings, ML-powered compliance tools are delivering measurable improvements in processing efficiency. For example, the integration of AI and machine learning in underwriting processes has led to a reduction in processing times by up to 50%, minimizing errors and enhancing scalability [7]. This efficiency gain translates directly to improved customer experience and increased lending capacity without proportional increases in compliance staff.
In the sanctions screening domain, AI-powered automation has shown remarkable efficiency improvements. According to recent industry analysis, 99.6% of sanctions screening alerts in some financial institutions are false positives, creating friction, fatigue, and risk to customer satisfaction. AI-powered automation built into alert investigation platforms can streamline case management, shorten resolution times, and reduce the conversion time to regulatory violations.
Lenders seeking to navigate the complex regulatory environment of 2025 must integrate ML-powered compliance tools throughout the entire lending lifecycle, from loan origination and credit assessment to servicing and recovery. This holistic approach creates a seamless compliance framework that adapts to changing regulatory requirements whilst maintaining operational efficiency.
African Bank provides an instructive example of this approach. The bank is launching a digital lending business that will streamline the entire lending process, using tools for credit assessment, loan approval, disbursement and recovery [6]. By embedding ML-powered compliance tools at each stage of the lending process, financial institutions can create a comprehensive compliance framework that adapts to changing regulatory requirements whilst maintaining operational efficiency.
The integration of advanced technologies throughout the entire lending lifecycle represents a holistic approach to compliance that goes beyond point solutions. Rather than addressing compliance requirements in silos, this approach ensures that compliance considerations are embedded in every aspect of the lending process.
The transition to cloud-based loan origination systems is accelerating this integration, offering scalability, enhanced security, and seamless compliance capabilities. This trend is expected to continue, with more lenders adopting these systems to improve digital loan processing while maintaining regulatory compliance [8].
Digital lending and eClosings are becoming increasingly prevalent, with Loan Origination Systems facilitating digital and mobile services that allow borrowers to apply for and manage loans via websites and mobile apps. This trend includes "eClosings," where documents are signed electronically, expediting the loan completion process while maintaining regulatory compliance [11].
Provira's experience illustrates the benefits of this approach: "Thanks to LendFusion, we've streamlined our operations and now provide faster service to our clients," noted their CEO after implementing automated repayment tracking and data-driven reporting for better control and precision [14]. The company needed a loan management system capable of handling complex financial transactions efficiently to help beneficiaries access inheritance funds. By adopting automated repayment tracking and data-driven reporting, they achieved better control and precision in their operations while ensuring compliance with relevant regulations.
Planet42, which provides car subscription services in South Africa, successfully expanded its reach without additional administrative staff by integrating automated loan origination, servicing, and multi-currency support. This enabled faster decision-making and improved customer satisfaction due to quicker application processing times, all while maintaining compliance with local regulations [14].
kennek's end-to-end credit platform exemplifies this approach, providing a unified system that supports compliance across the entire lending lifecycle. By centralising data and processes, kennek reduces the risk of compliance gaps that can occur when using fragmented systems. This centralised approach also enables real-time monitoring and reporting, allowing lenders to identify and address potential compliance issues before they become regulatory violations.
The key to successful implementation lies in selecting a platform that is both comprehensive and flexible. The platform should cover all aspects of the lending lifecycle whilst allowing for customisation to meet the specific regulatory requirements of different jurisdictions. It should also integrate seamlessly with existing systems and workflows to minimise disruption and maximise adoption.
As we look beyond 2025, it's clear that regulatory requirements will continue to evolve. Forward-thinking lenders are developing adaptable compliance frameworks that can evolve alongside these requirements, including the use of ML algorithms that can identify emerging regulatory trends and automatically adjust compliance processes to address new requirements before they become mandated.
The future of financial services will be shaped by those who use compliance as a catalyst for innovation and growth, rather than just compliance [5]. This perspective represents a fundamental shift in how lenders approach regulatory requirements—from a reactive necessity to a proactive strategic advantage.
By leveraging ML algorithms that can analyse regulatory trends and anticipate future requirements, institutions are developing compliance frameworks that adapt to evolving regulations before they become mandated. This proactive approach reduces compliance risk whilst providing first-mover advantages in an increasingly competitive market.
The rapid proliferation of fintech products, embedded finance solutions, and complex bank partnerships is further blurring the lines between lenders and non-lenders, creating additional regulatory complexity [9]. In this environment, regular dialogue with legislators, regulators, and participation in public policy discussions is becoming essential for lenders to address misconceptions about the industry and help shape balanced regulations.
The financial landscape is also witnessing significant structural shifts. Over the past decade, U.S. banks have significantly increased lending to Non-Bank Financial Institutions (NBFIs)—such as private credit funds, insurers, asset managers, and hedge funds—while direct commercial and industrial loans have stagnated. Bank lending to NBFIs has quintupled to over $1 trillion, now comprising more than 10% of total bank loans [15]. This shift is driven by regulatory capital efficiencies, as NBFI loans often carry lower risk weights due to strong collateral and structural protections. However, it also creates new regulatory challenges that require sophisticated compliance frameworks.
Additionally, fintech companies like Klarna, SoFi, and Upstart have been selling substantial volumes of consumer, auto, and home equity loans to private credit firms such as Elliott, Carlyle, Fortress, and Blue Owl. This trend allows fintechs to focus on their brands and underwriting algorithms while offloading balance sheet risks [16], but creates complex regulatory considerations that span multiple entities.
The rise of Synthetic Risk Transfers (SRTs) provides another example of evolving financial structures that create new compliance challenges. Banks have increasingly utilized SRTs to manage credit risk on their balance sheets. While traditionally more common in Europe, the U.S. market has seen significant growth in SRT activity, providing banks with efficient means to access high-quality bank-originated credit [10]. These complex financial instruments require sophisticated compliance frameworks that can adapt to evolving regulatory requirements.
Looking ahead, the global RegTech market, valued at approximately £10.26 billion in 2023, is projected to exceed £68 billion by 2032, growing at a compound annual growth rate (CAGR) of 23.6% [19]. This substantial growth reflects the increasing importance of regulatory technology in helping financial institutions navigate complex compliance requirements efficiently and effectively.
kennek's platform is designed with this future-proofing in mind. Its flexible, API-first architecture allows for rapid adaptation to changing regulatory requirements, whilst its real-time monitoring capabilities enable proactive compliance management. By centralising data and processes, kennek provides a single source of truth for compliance reporting, reducing the risk of inconsistencies and errors that can lead to regulatory violations.
As regulatory requirements continue to evolve, the ability to adapt quickly and efficiently will become an increasingly important competitive advantage. Lenders that invest in adaptable compliance frameworks today will be better positioned to navigate the regulatory challenges of tomorrow, turning what many see as a burden into an opportunity for differentiation and growth.
The regulatory environment for lenders in 2025 is characterised by fragmentation and complexity, with state regulators stepping up to fill the void left by diminished federal oversight. This environment creates significant challenges for lenders operating across multiple jurisdictions, each with its own enforcement priorities and compliance requirements.
ML-powered compliance tools offer a solution to these challenges, enabling lenders to navigate the complex regulatory environment more efficiently and effectively. By automating routine compliance tasks, reducing false positives in fraud detection, and providing predictive insights into potential compliance issues, these tools transform compliance from a cost centre to a strategic advantage.
The integration of ML-powered compliance tools throughout the entire lending lifecycle creates a seamless compliance framework that adapts to changing regulatory requirements whilst maintaining operational efficiency. This holistic approach ensures that compliance considerations are embedded in every aspect of the lending process, reducing the risk of compliance gaps and enabling proactive compliance management.
As regulatory requirements continue to evolve, the ability to adapt quickly and efficiently will become an increasingly important competitive advantage. Lenders that invest in adaptable compliance frameworks today will be better positioned to navigate the regulatory challenges of tomorrow, turning what many see as a burden into an opportunity for differentiation and growth.
kennek's end-to-end credit platform exemplifies this approach, providing a unified system that supports compliance across the entire lending lifecycle. By centralising data and processes, kennek reduces the risk of compliance gaps that can occur when using fragmented systems, enabling lenders to navigate the complex regulatory environment of 2025 and beyond with confidence.
The current regulatory environment, marked by increasing fragmentation and shifting priorities, presents significant operational challenges for non-bank lenders. We see clearly that relying on disparate, legacy systems is no longer sustainable. The rising tide of delinquencies further underscores the critical need for sophisticated, data-driven approaches to risk and compliance. Our analysis shows that AI and machine learning are not merely tools but fundamental requirements for navigating this complexity effectively. They enable the automation necessary to manage diverse jurisdictional requirements and transform compliance from a burdensome cost centre into a source of measurable efficiency and predictive insight.
To truly leverage these capabilities and achieve strategic advantage, lenders require an integrated, end-to-end platform that centralises data and workflows across the entire lending lifecycle. This holistic approach ensures compliance is embedded from origination through to servicing and recovery, providing the adaptability needed to future-proof operations against evolving regulations. We believe this integrated framework is the only viable path for lenders to scale confidently, enhance decision-making, and maintain robust compliance while driving growth in a dynamic market.
Xavier De Pauw is the co-founder & CFO of kennek, a complete lending software for alternative credit. A seasoned banker turned fintech entrepreneur, Xavier spent 10 years at Merrill Lynch specialising in structured finance before co-founding challenger banks MeDirect Group and MeDirect Bank Belgium, building a €2.5 billion balance sheet and acquiring 32,000 retail clients. He also co-founded Fintech Belgium, serving as president to foster fintech growth. Most recently, he led Strategic Innovation & Marketing at the private bank, Degroof Petercam, driving impactful digital transformations.
[1] State Authorities Stepping Up Consumer Protection Efforts, Consumer Finance Monitor
[2] State Enforcement, Consumer Finance Monitor
[3] Monthly Bank Regulatory Report - January 2025, Gibson Dunn
[4] Lending Redefined: Implementing ML To Enhance Products, Workflow, and Customer Experience, Finextra
[5] Our Mission Is To Deliver Scalable, Secure, And Globally Benchmarked Platforms Deeply Relevant To Modern Financial Ecosystems, Digital Terminal
[6] African Bank readies digital lending business for SMEs, ITWeb
[7] The Future of Credit Underwriting: Trends and Technologies to Watch, Writer Information
[8] 10 Loan Origination Trends to Watch in 2025, LendFoundry
[9] Alternative Commercial Finance Monthly - May 2025: Things Heard at a Factoring Conference, Husch Blackwell
[10] Regulatory Challenges with Digital Banks and Alternative Lenders, arXiv
[13] Transforming Loan Management: Case Study on Alta West Capital's Experience, Fundingo
[14] Case Studies: Loan Management Software, LendFusion
[15] Growth of Non-Bank Financial Institutions, Financial Times
[16] Fintechs Selling Loans to Private Credit Firms, Financial Times
[17] Why Financial Institutions Are Turning to AML Overlays, Not Replacements, Fintech Global
[18] AI and ML in Regulatory Compliance, FlipHTML5
[19] Top 10 RegTech Trends for 2025, Bobsguide
[20] Global Consumer Lending Confidence Report, LexisNexis Risk Solutions
[21] Gartner Survey Shows 58 Percent of Finance Functions Use AI in 2024, Gartner
[22] AI-Driven Risk Analysis: The New Standard in Financial Strategy, Tech Times
[23] Financial IT Summer Edition 2025, Financial IT
[25] Revolutionizing KYC Leveraging AI/ML for Regulatory Compliance, DZone