New Relic Research: 22% of Financial Firms Deploy AI Agents Without Critical Monitoring

Financial services firms are rapidly deploying AI agents into production environments, yet nearly a quarter are doing so without any monitoring in place. This visibility gap creates significant operational risks and potential compliance failures, particularly as regulations like the Digital Operational Resilience Act (DORA) tighten requirements for digital resilience across the global banking and insurance sectors.

What was announced

New Relic’s 2026 Observability Forecast for Financial Services highlights a disconnect between AI adoption and system oversight. The study, which surveyed 278 industry leaders as part of a larger global cohort of 2,575 respondents, found that 22% of financial organisations have live AI agents in production with zero monitoring. This lack of oversight introduces what the report calls "phantom velocity," where AI-driven productivity gains are negated by unmonitored downtime and emergency fixes caused by the systems themselves.

The report notes that while the sector currently trails the global average in AI application monitoring—42% versus 47% globally—it is poised for a rapid catch-up. Within the next 12 months, AI monitoring adoption in financial services is projected to reach 83%, potentially overtaking the global trajectory of 80%. This shift is driven by the recognition that AI-generated code requires more rigorous observability; 78% of industry leaders agree that such code makes reliable monitoring more critical.

Operationally, the sector remains resilient compared to other industries. Financial firms experience high-impact outages at a lower rate, with 30% reporting weekly occurrences compared to 36% in the broader market. When failures occur, they are resolved faster, taking an average of 49 minutes compared to the global 54-minute average. The financial impact is also lower, with outages costing a mean of $1.69 million per hour against a global average of $1.85 million.

The companies involved

New Relic is an observability platform designed to help engineers and IT professionals monitor, debug, and improve their software stacks. The company provides a data-driven approach to software engineering, allowing organisations to collect telemetry data from various sources to gain a comprehensive view of their digital operations. In the context of financial services, the platform is used to manage complex legacy systems alongside modern cloud-native applications and emerging AI technologies.

The company occupies a critical position in the market as financial institutions face increasing pressure to modernise their infrastructure while adhering to strict regulatory frameworks. By providing visibility into application performance and infrastructure health, the platform aims to reduce mean time to resolution (MTTR) and prevent costly outages. As the industry moves toward agentic AI—systems capable of making autonomous decisions—the role of observability platforms becomes central to maintaining the operational integrity and compliance standards required by global financial regulators.

What this means

The move toward agentic systems represents a fundamental shift in financial technology, moving from static automation to autonomous decision-making. However, the 22% monitoring gap suggests that many firms are prioritising speed over safety, a dangerous gamble in an era of strict DORA enforcement. If an unmonitored agent triggers a flash crash or a series of erroneous transactions, the regulatory fallout will be severe. The industry is currently under pressure to prove that AI is a controlled asset rather than a "black box" liability. The real test for the sector will be whether monitoring tools can evolve fast enough to interpret the non-linear logic of AI agents.