As financial institutions introduce AI agents that can interact with customers and carry out tasks, proving these systems deliver appropriate outcomes is becoming a growing compliance challenge. The AI in Risk and Compliance 2026 report, co-authored by RegTech Analyst and Parker & Lawrence Research, explored how firms can strengthen oversight as AI takes on a greater role in regulated financial services.
As part of the research for the report, Parker & Lawrence Research interviewed providers developing solutions to oversee AI systems operating directly with customers in regulated markets.
On this occasion, the firm spoke with Aveni, a provider of AI-powered oversight and assurance for customer interactions in regulated financial services.
This interview was part of the wider AI in Risk and Compliance 2026 report conducted by RegTech Analyst and Parker & Lawrence Research. To download the full report, click here.
Traditional AI testing can establish whether a model meets technical benchmarks or withstands defined failures, but it cannot necessarily demonstrate how a system will behave during a real customer interaction. An AI agent may perform well in controlled conditions while failing to recognise vulnerability, respond appropriately to a customer’s circumstances or deliver a suitable outcome. For regulated firms, technical accuracy is only part of the challenge.
The issue is becoming more pressing as AI deployment accelerates. Research cited in the report found that 65.6% of reported AI activity had reached production or beyond, while 69.3% of institutions said deployment had outpaced their measured control environment. Existing model risk and quality assurance frameworks can struggle to accommodate AI agents conducting lengthy, dynamic conversations. Pre-deployment testing may overlook unusual combinations of language and customer circumstances, while manual reviews and sample-based checks can leave serious but infrequent conduct risks undetected.
Financial institutions must also balance the depth of oversight against operational costs and response times. Running resource-intensive models on every conversational turn can increase latency and expense, potentially undermining the customer experience. Firms therefore need ways to monitor interactions continuously without creating new operational bottlenecks or weakening existing risk and compliance controls.
Aveni aims to address this challenge through a three-part approach covering AI testing before deployment, real-time monitoring and post-interaction quality assurance. Its Approve product connects to customer-facing AI through an API, allowing firms to simulate conversations using selected customer personas and risk scenarios. These dialogues are assessed against criteria including vulnerability, advice boundaries, harmful language, customer understanding and escalation, giving model and risk committees evidence of what passed or failed and why. The process can be repeated whenever a model changes.
Once an AI system is live, Aveni’s Assure product monitors conversations through a real-time conduct risk guardrail. Faster, task-specific models carry out initial checks within the available latency budget, while higher-risk exchanges can receive more detailed analysis. Depending on the circumstances, the system can allow a conversation to continue, alert a human reviewer, redact content or stop the interaction.
Aveni’s Detect product focuses on post-interaction quality assurance across voice, chat, email, documents and related records. Rather than relying solely on samples, it applies a firm’s own quality assurance and Consumer Duty framework across entire cases, retaining evidence for review and remediation. The product can also identify potential upselling opportunities and early indications of customer churn.
According to Aveni, one quality assurance use case reduced assessment time from 90 minutes to 15 minutes. The company also reported 61% faster time to first token, a 3% improvement in F1 accuracy and 92% lower running costs than Gemini Flash 2.5 Lite in benchmarked use cases. These figures were supplied by Aveni and relate to its specific tests.
Aveni chief commercial officer Kent Mackenzie said, “The bank that solves the oversight problem will be the bank that wins the race to become the most agentic bank.”
Parker & Lawrence Research highlights Aveni’s combination of customer outcome-focused oversight and specialist engineering. The company has trained its own small language models on financial services data and orchestrates more than 150 models, ranging from low-latency classifiers to specialist financial services models and selectively deployed frontier models.
The analysts argue that making testing across entire populations of customer interactions more practical could help firms identify risks that manual sampling might miss. As banks introduce increasingly autonomous AI, the ability to demonstrate appropriate customer outcomes may become as important as technical performance, with effective oversight helping firms build confidence in deploying these systems.
Read the original post from Parker & Lawrence Research here.
Copyright © 2026 FinTech Global
Copyright © 2026 RegTech Analyst
