Compliance Pressure Is Building in Digital Collections and AI Deployment: What Needs to Change Before Regulators Act

By Nick Cherry, Divisional CEO, Phillips & Cohen Associates, Ltd.

I spend a lot of time talking with credit and collections leaders who are in the middle of expanding their digital engagement strategies, often through the implementation of agentic AI technology. For most organizations, that shift is no longer optional. Consumers increasingly prefer text, email, and secure portals over phone calls and letters. Digital outreach is faster, more efficient, and often more aligned with how people manage their finances today.

On the surface, that modernization looks like progress. Engagement rates improve. Operational costs stabilize. Consumers respond through channels they are comfortable using at a time and place that suits them. Underneath that progress, however, I see a quieter tension building.

The compliance frameworks that govern collections were largely designed for a different communication environment. They were built around call monitoring, script approvals, and structured documentation tied to voice and written correspondence. Digital engagement scales quickly and expands communication volume in ways traditional frameworks were not built to monitor. AI-driven processes add another layer, operating through automated decision logic that is not always fully visible across the organization.

When those capabilities expand faster than oversight structures evolve, subtle gaps begin to form. Those gaps rarely appear in standard performance dashboards. They show up in complaint patterns, audit findings, model performance issues, or regulatory reviews.

The Forces Elevating Digital and AI Compliance Risk

The pressure is not coming from one direction. It is cumulative. Delinquency volumes are rising across many portfolios, which naturally increases total outreach attempts. At the same time, AI-driven digital channels allow for more tailored touchpoints and engagement strategies. 

AI tools are also becoming more embedded across collections operations. Some influence outreach cadence or segmentation. Others sit behind the scenes, supporting agent efficiency, quality improvement, account prioritization, workflow management, or operational decision making. These tools can improve efficiency and consistency, but they also increase the importance of understanding how decisions are being made and monitored.

Increased touchpoints and automation mean more opportunities for inconsistency. I also see greater visibility into communication patterns. Consumers can easily screenshot or forward digital messages. Small variations in language or timing that might once have gone unnoticed can now circulate widely. That does not mean digital outreach or AI-driven processes are inappropriate. In many cases, they improve clarity and efficiency. But they do require consistent governance.

Regulators are paying attention to this shift as well. The expectation is not that organizations avoid digital engagement or AI. In actuality, it is that they demonstrate control over these systems and ensure that they remain accurate, proportionate, and appropriately monitored.  Well-deployed and governed agentic AI tools can transform an organisation’s efficiency levels and operational capability, but their use also amplifies existing risks if adequate controls & oversight are not established to govern them. 

Where Digital and AI Collections Create Blind Spots

In conversations with lenders and agencies, the challenges tend to be operational rather than intentional.

Consent management is one example. Many organizations capture consent appropriately at origination or early servicing stages. The difficulty comes later, when accounts transfer between systems or partners. If consent documentation is not consistently maintained and accessible, it becomes harder to demonstrate compliance across the full lifecycle of the account.

Record retention is another area that requires attention. Call recordings typically sit within established storage systems. Digital interactions may occur across multiple platforms, each with different archival practices. When automated systems influence outreach timing or account treatment, organizations must also be able to explain why a particular action occurred. Without that visibility, reconstructing a full communication history can become more complicated than expected.

Automation adds another layer. Workflow triggers, payment reminders, and templated messaging can operate efficiently at scale. AI-driven decisioning can also shape prioritization and segmentation behind the scenes. But automation does not eliminate the need for review. In fact, it increases the importance of periodically validating that message logic, system behavior, and decision outcomes remain aligned with policy and regulatory expectations.

I also see many firms still building visibility into where AI is being used across the business. Different teams may adopt tools independently. Over time, that can make it difficult to maintain a clear view of how automated decisioning is operating across consumer outreach and back-office functions.

Data fragmentation between creditors and agencies can introduce additional risk. If hardship status, disputes, language and channel preferences, or other key data points are not shared in near real time, outreach may inadvertently conflict with a consumer’s current situation. AI systems are only as reliable as the data they receive. Inconsistent or incomplete data inputs will affect outcomes.

Strengthening Governance in a Digital and AI Environment

The solution is not to slow digital or AI expansion. It is to strengthen the governance that supports it.

First, organizations should align channel growth and technology deployment with formal compliance reviews. Every expansion of digital capability or AI use should prompt an assessment of consent documentation, monitoring processes, system logic, and audit trails. Data governance and cyber risk should be considered with every new deployment. Growth and oversight must move together.

Second, quality assurance models must evolve. Reviewing a small sample of calls is not equivalent to overseeing thousands of automated messages or AI-influenced decision points. Monitoring frameworks should include testing of workflow logic, template updates, system outputs, and performance over time.  

Third, leadership teams should be clear about ownership and accountability. It should be evident who is responsible for reviewing system performance, monitoring outcomes, and escalating issues when they arise. As automation expands, clarity around responsibility becomes more important, not less.

Finally, data alignment across the credit ecosystem needs to improve. Near real-time sharing of disputes, hardship indicators, and consumer preferences reduces the likelihood of inconsistent outreach. When lenders and agencies operate from the same current information, both compliance posture and consumer experience improve.

Acting Before Scrutiny Forces Change

Regulatory cycles tend to follow growth cycles. As digital engagement and AI deployment become more central to collections operations, oversight expectations will continue to rise. Because policies and practices around AI use are still evolving across the industry, there is an opportunity for firms to establish strong governance before scrutiny forces rapid change.

The firms that will be best positioned over the next several years are those that treat digital modernization, AI deployment, and compliance modernization as parallel efforts. Digital collections and AI are here to stay. The question is whether our control structures evolve alongside them.