Lloyds Banking Group is accelerating its shift from controlled artificial intelligence pilots to full-scale deployment inside core engineering workflows, as a new research partnership with the University of Glasgow highlights how testing, governance and delivery are converging in modern banking.
The four-year programme will embed agentic AI systems directly into day-to-day software and data engineering, with teams working alongside large language model-based coding agents across multiple global hubs.
The initiative is designed to measure real-world performance and establish best practices for scaling AI safely across a large, regulated organisation.
For QA and software testing teams, this marks a clear departure from traditional validation models. AI is no longer something tested in isolation before release. It is increasingly something that must be continuously validated inside live engineering environments, under real constraints and risks.
“Agentic-driven software engineering is a fast-developing sector with the potential to enable human engineers to work more efficiently by automating some tasks and allowing them to focus their skills on higher-level work,” explained Dr Tim Storer of the University of Glasgow.
However, Storer was keen to stress “there has been relatively little research in industry on how integrating agentic AI into software engineering practices can be done effectively in large-scale organisations.”
He added that the aim is to enable the banking group’s plans “to increase their software development capacity, produce high-quality research for the benefit of all, and influence national policy and industry standards.”
The programme builds on Lloyds’ earlier work inside its innovation sandbox with UnlikelyAI, where the bank tested neurosymbolic AI systems designed to improve transparency and reliability.
That initiative aimed “to explore how neurosymbolic AI can be used across the Group, in a way that delivers responsible, accurate and consistent support,” with systems evaluated alongside existing AI platforms.
The approach, combining neural networks with symbolic reasoning, is intended to produce outputs that are explainable and less prone to hallucination.
For QA teams, this introduces a more complex assurance challenge, requiring validation not just of outputs, but of reasoning, consistency and explainability.
Ranil Boteju, the London-based Chief Data and Analytics Officer at Lloyds Banking Group, stressed that the bank is “harnessing the power of AI to revolutionise banking.”
He stressed that “UnlikelyAI’s neurosymbolic approach allowed us to test innovative new features for cthat are not only useful, fast and intelligent, but also transparent and reliable.”
Boteju also added that the trial was designed to understand “how cutting-edge artificial intelligence can support innovation and enhance customer experience” while maintaining the auditability and explainability required in regulated environments.
Regulators push banks
Lloyds’ expansion of AI testing into engineering workflows comes as the Financial Conduct Authority steps up pressure on banks to prove how AI behaves in live conditions.
Through its AI Live Testing programme, which includes firms such as Barclays and UBS, the regulator is pushing institutions beyond experimentation and into supervised, real-world validation.
Jessica Rusu, the FCA’s chief data, information and intelligence officer, explained recently that “we’re continuing to collaborate with firms to support the safe and responsible development of AI in UK financial markets.”
“With tailored support from the FCA and Advai, the initiative reflects our commitment to supporting the pace of change in AI, whilst demonstrating how regulators and industry can work together to harness innovation responsibly,” she added.
For QA teams, the implication is clear. Testing is no longer a downstream activity. It is becoming a regulatory artefact, expected to generate evidence that AI systems behave safely, remain controlled and perform reliably under real operational conditions.
Sustainability pressures
At the same time, Lloyds’ AI push is intersecting with another emerging constraint for testing teams: sustainability.
Speaking at the QA Financial Forum London 2025, Richard Bishop, Lead Quality Engineer at Lloyds Banking Group, warned that the expansion of automated testing and AI-driven workloads could significantly increase the environmental footprint of software delivery.
“Software testing environments produce 120% more carbon emissions than production environments,” the London-based industry veteran told conference delegates.
He added that the broader context is already significant: “The IT sector is responsible for 2-3% of global emissions, which is comparable to the airline industry.”
As banks scale data-intensive AI and automated testing, this creates a new balancing act for QA teams, who must now optimise not only for quality, speed and compliance, but also for energy consumption and emissions.
Bishop said at the industry event that “yes, there is a significant risk that increased automation and especially the use of data-intensive GenAI tools could lead to higher emissions.”
He stressed “automation and AI require significant computational power, which can increase energy consumption. However, this can be mitigated by optimising the efficiency of these tools and processes.”
Embedding governance, quality and carbon
Lloyds has begun aligning its quality engineering practices with sustainability goals, introducing dashboards and metrics to track emissions across testing environments and projects.
This reflects a broader shift in how banks are structuring software delivery, where testing, governance and sustainability are no longer separate concerns.
Bishop said: “By aligning quality benchmarks with CO2 reduction targets, we hope to create a more sustainable software development lifecycle.”
He added that embedding sustainability across the lifecycle is critical: “Focusing on ‘sustainability in projects’ rather than just ‘sustainability in testing’ broadens the scope of environmental responsibility to remove the idea that sustainability is ‘someone else’s problem’.”
This mirrors the direction of Lloyds’ AI research programme, which is explicitly designed to generate “rigorous, real-world evidence” from engineering workflows.
Dr Shane Montague, Head of Research Engineering at Lloyds Banking Group, explained that “our mission to means leading innovation that genuinely improves how engineering gets done, with a focus on delivering enhanced digital services.”
He said the partnership with the University of Glasgow is vital to gather rigorous, real-world evidence from day-to-day engineering work, so “we can understand what really works and how agentic AI can be applied effectively and responsibly at scale.”
Taken together, Lloyds’ initiatives underline how quickly the role of QA is evolving inside banks. AI systems are moving into core engineering processes. Regulators are demanding real-world validation. And sustainability is emerging as a measurable constraint on testing at scale.
For QA and software testing teams, the result is a fundamental shift. Assurance is no longer about verifying whether systems work. It is about proving that they behave correctly, remain explainable, meet governance standards and do so efficiently enough to be sustainable in production banking environments.
THIS MONTH
Why not become a QA Financial subscriber?
It’s entirely FREE
* Receive our weekly newsletter every Wednesday * Get priority invitations to our Forum events *
REGULATION & COMPLIANCE
Looking for more news on regulations and compliance requirements driving developments in software quality engineering at financial firms? Visit our dedicated Regulation & Compliance page here.
READ MORE
- Inside banking’s shift to smarter QA to tackle complexity and risk
- SmartBear CPTO on AI in banking QA: ‘Impressive metrics but no critical scenarios’
- Banks push beyond traditional QA as resilience testing gains ground
- Banking QA professionals warn AI still doesn’t know ‘where the bodies are buried’
- RECAP: The QA Financial Healthcare & Insurance Forum Philadelphia 2026
WATCH NOW
QA FINANCIAL PODCASTS
CLICK HERE TO LISTEN TO OUR EXCLUSIVE CONVERSATIONS
