AI Adoption Outpaces Governance in Financial Services as 64% of Firms Admit to Releasing Untested Code

Quick Summary

New research reveals that AI governance readiness is lagging behind rapid adoption in financial services. While 70% of firms use AI in software delivery, 64% admit to releasing untested code. This gap creates significant operational resilience risks, with poor software quality costing some firms over £744,000 annually.

How Does AI Adoption Impact Software Quality in Banking?

The rapid integration of AI-driven development is creating a volume of code that traditional testing teams cannot manage. According to the Tricentis 2026 Quality Transformation Report, 30% of financial services professionals state that AI generates more code than can be realistically tested. This surge in output, combined with leadership pressure for speed, has led to a culture where 64% of global firms are deploying software without completing necessary quality checks.

• 64% of firms admit to releasing untested software code.
• 30% of teams feel pressured by leadership to bypass testing.
• 27% of organisations cite security breaches as the top risk of poor quality.

What Are the Financial and Regulatory Risks of Poor AI Governance?

Failing to establish robust AI governance readiness carries a heavy price tag for the financial sector. Nearly one in five organisations estimate that poor software quality costs them more than £744,000 every year. Beyond direct costs, the regulatory compliance failures and loss of customer trust are viewed as existential threats. As regulators increase oversight on autonomous systems, the 60% of firms who feel unprepared to govern AI agents face a significant compliance gap risk.

• £744,000+ annual cost for 18% of surveyed financial firms.
• Only 40% of firms feel prepared to govern and scale AI agents.
• 80% trust AI to make release decisions despite governance gaps.

How Can Firms Align AI Innovation with Operational Resilience?

Bridging the gap requires a shift toward agentic quality engineering that matches the speed of development. Currently, only 38% of developers and executives are aligned on release readiness, suggesting a breakdown in internal communication. To maintain operational resilience, firms must prioritize AI-powered testing that automates oversight. Andrew Power of Tricentis notes that AI governance readiness must become a board-level priority, similar to how cybersecurity evolved over the last decade.

FF NEWS TAKE:

This report is a wake-up call for the industry. While AI governance readiness might sound like a back-office concern, the fact that 64% of firms are shipping untested code in a highly regulated environment is staggering. This moves the needle by highlighting that the next big fintech crisis won't be a market crash, but a software quality failure triggered by unmanaged AI. Firms must stop choosing between speed and safety.