AI voice technology is gaining interest across the accounts receivable management industry, but adoption is running into a less technical obstacle: workforce resistance. A new field report from Vodex, developer of the DROS AI-native collections platform, argues that collections leaders are increasingly comfortable discussing compliance and voice quality, yet remain cautious about deploying AI that employees could perceive as a threat to their jobs.
The collections industry may be closer to adopting AI voice technology than deployment numbers suggest. The obstacle, according to new field research from Vodex, is increasingly not whether the technology can work, but whether collections organizations are comfortable with what it could mean for their human workforce.
Vodex’s report, “The Adoption Gap: State of AI in Collections 2026,” draws on more than 25 open-ended conversations with agency owners, compliance leaders and operations managers at the ACA International Annual Convention & Expo in Orlando in July. The company describes the research as qualitative and directional rather than statistically representative, since participants were self-selecting attendees who visited the DROS booth.
The report’s central observation is that compliance tends to be the first objection raised when collections operators evaluate AI voice, but workforce concerns often emerge as the more consequential barrier.
That distinction could matter for vendors selling conversational AI into accounts receivable management. Operators may initially ask whether an AI agent can handle requirements surrounding the Fair Debt Collection Practices Act (FDCPA), Regulation F, disclosures and consent. But once those questions are addressed, the harder question becomes what happens to the human collections floor.
Vodex says many operators it interviewed were reluctant to introduce technology that their strongest collectors might interpret as a threat to their roles. The company argues that this creates a messaging problem for the industry: AI is frequently positioned around reducing headcount or lowering labor costs when a more compelling use case may be expanding contact coverage.
Instead of replacing collectors, AI voice could be used to reach accounts that existing teams lack the capacity to contact consistently. In that model, an automated voice channel becomes an additional layer of collection capacity rather than a direct substitute for human agents.
That distinction mirrors a broader challenge facing enterprise AI. McKinsey’s 2025 global AI survey found that 62% of respondents said their organizations were experimenting with AI agents, while 23% reported scaling an agentic AI system somewhere in the enterprise. Yet no more than 10% reported scaling agents within any individual business function.
The collections industry has another problem: technology perceptions can lag behind the current generation of products.
Vodex says several operators it interviewed had evaluated conversational AI two or three years ago and rejected it because of unnatural speech or excessive latency. The company argues that improvements in voice quality and response times have created a different technical baseline, but some buyers are still judging today’s systems against earlier-generation experiences.
For a highly conversational process such as debt collection, that perception can be important. A voice system that sounds artificial or struggles with interruptions can undermine consumer trust and reduce the likelihood that operators will consider it suitable for production.
Compliance presents a more complicated picture. Debt collection is governed by detailed rules concerning communications, disclosures, prohibited conduct and recordkeeping. The Consumer Financial Protection Bureau’s Regulation F implements the FDCPA and includes requirements covering communications, validation information, harassment and abuse, misleading representations, time-barred debts and record retention.
Vodex argues that these requirements can actually strengthen the case for carefully engineered AI voice. A software-controlled agent can theoretically deliver a required disclosure consistently, enforce predefined conversational rules and generate a complete recording and transcript for review.
That does not eliminate compliance risk. It changes where the risk sits.
A human collector can make an inconsistent disclosure or depart from policy; an automated system can repeat the same mistake at scale if its logic is incorrectly designed. For that reason, versioned scripts, policy controls, monitoring, escalation mechanisms and comprehensive records become important parts of an AI collections architecture.
The report identifies inbound overflow and after-hours coverage as relatively low-resistance starting points for organizations considering an initial deployment. That approach also provides a narrower environment in which operators can measure contact rates, resolution outcomes, escalation frequency, consumer experience and compliance performance before expanding AI into more sensitive workflows.
Another shift could come from distribution. Vodex says platform and system-of-record vendors are increasingly embedding AI directly into their products. If that trend accelerates, collections agencies may increasingly acquire AI capabilities through existing software relationships rather than selecting a separate conversational AI provider.
For collections technology vendors, that creates a competitive challenge. AI voice may become less differentiated as a standalone feature and more dependent on integration with collection-management systems, customer data, compliance controls and workflow engines.
For HR leaders, the development is equally significant. Collections organizations adopting AI will need workforce strategies that explain how technology changes roles rather than simply promising labor savings. Human collectors could increasingly focus on complex negotiations, escalations and cases requiring judgment while automated systems handle repetitive outreach and coverage gaps.
The result may be less about replacing the collections floor and more about restructuring it.
Vodex’s research should not be interpreted as proof that the entire ARM industry shares the same concerns. Its sample is deliberately small and self-selected. But the report offers a useful signal: AI adoption can stall even when technology and compliance questions have plausible answers if employees and managers do not see a credible role for humans in the resulting operating model.
For AI voice to move deeper into collections, vendors may therefore need to sell more than automation. They will need to demonstrate measurable coverage gains, reliable compliance controls and a workforce model in which AI expands what human teams can accomplish.
Market Landscape
AI adoption in collections sits at the intersection of conversational AI, fintech infrastructure, compliance technology and workforce automation. The market is shifting from experimental chatbots toward AI agents capable of completing defined workflows, but enterprise-scale adoption remains limited.
McKinsey reports that 62% of organizations are experimenting with AI agents and 23% are scaling agents somewhere in the enterprise, while fewer than 10% report scaling them in any individual function.
In debt collection, regulatory requirements add another layer. Regulation F establishes federal requirements governing debt-collection communications and prohibits conduct including harassment, abuse and false or misleading representations.
The competitive opportunity is consequently moving beyond voice synthesis. Vendors will compete on compliance architecture, integrations, analytics, workflow automation, human escalation and measurable collection outcomes.
Top Insights
- AI voice adoption in collections may be constrained more by workforce concerns than by technical readiness or initial compliance questions.
- Positioning AI as additional contact capacity could prove more acceptable than presenting automation primarily as a mechanism for reducing headcount.
- Collections organizations need auditable AI systems because automated compliance errors can potentially be repeated at scale.
- Inbound overflow and after-hours coverage offer narrower starting points for testing AI voice before expanding into core collection workflows.
- Embedded AI from collections platforms and system-of-record vendors could change how agencies acquire and deploy conversational automation.
Join thousands of HR leaders who rely on HRTechEdge for the latest in workforce technology, AI-driven HR solutions, and strategic insights
