A new McKinsey book excerpt is shedding light on a shift that could fundamentally reshape how banks design, test and validate software, as AI moves from assisting developers to actively executing large parts of the development and testing lifecycle.
Inside financial institutions, this transition is no longer theoretical. The model described points to environments where software delivery, validation and risk assessment are happening continuously, with AI systems generating not just code, but also the evidence, test coverage and control signals required to support it.
For QA and software testing teams, the implications go well beyond productivity. The shift redefines how quality is engineered, how risk is identified, and how resilience is maintained in increasingly complex, AI-driven systems.
“If gen AI has a killer application, it’s software development, one of the most profound shifts in the history of programming,” according to the book, titled Rewired: McKinsey’s Playbook on How Leading Companies Win with Technology and AI.
That shift is already playing out inside at least one large global systemically important bank, where “AI agent teams, nearly a hundred of them, have just finished their shift, having spent the night refining a new cross-border payment system, testing failure paths, and shipping updates at a pace no human team could match.”
For testing teams in banking, the detail that stands out is not just the acceleration of development, but the embedding of testing into that process.
By morning, engineers are met with “a neatly organized stream of AI-generated pull requests, test evidence, and risk flags, more progress in 12 hours than a traditional team might make in a month,” according to the book.
Testing, in this model, is continuous, automated and inseparable from development. AI is not only generating code, but also producing the validation artefacts required to assess its quality and safety.
“If gen AI has a killer application, it’s software development, one of the most profound shifts in the history of programming.”
– McKinsey book
The traditional rhythm of software delivery is also being reworked. In place of fixed sprint cycles, the model points to a continuous loop of development, testing and validation.
“Software development becomes a continuous, high-speed loop rather than a two-week sprint cycle,” the authors of the book wrote.
AI agents are increasingly responsible for executing complex workflows across the lifecycle, including “creating evidence provenance, running legal and cyber checks, testing counterfactuals, and both suggesting and making decisions.”
For financial institutions, this introduces a new paradigm where testing expands beyond functional validation into areas such as compliance, security and scenario-based risk testing, all within the same automated loop.
The rise of the AI testing ‘factory’
At the centre of this transformation is the concept of an AI-driven development and testing factory, where multiple specialised agents handle different aspects of software delivery and validation.
In this model, “test agents generate and run new test suites; QA agents identify regressions; security agents scan for vulnerabilities or leaked secrets; performance agents benchmark critical paths.”
For banks, this introduces a form of industrialised testing at scale, where validation is continuous and deeply integrated into delivery pipelines. At the same time, it raises new questions around governance, traceability and the reliability of AI-generated test outcomes.
An orchestration layer coordinates these activities, ensuring that failures are addressed automatically. “If tests fail, it routes work back to a fix agent; if performance declines, it invokes a performance-checking agent; if a policy is violated, it halts the workflow.”
McKinsey’s analysis of nearly 300 companies suggests that the benefits of AI in software development are closely linked to how deeply it is embedded across the full lifecycle, including testing.
Top performers are achieving “16–30 percent improvements in productivity, time to market, and customer experience, along with 31–45 percent gains in software quality.”
However, the findings also highlight that tooling alone is insufficient. “Simply giving developers AI tools does not meaningfully move the needle.”
Instead, leading organisations “deploy multiple AI development use cases spanning ideation, requirements, design, coding, testing, deployment, and operations,” embedding testing and quality assurance into every stage.
For QA teams, this reinforces the need to evolve alongside development practices, ensuring that validation frameworks can keep pace with AI-driven delivery.
Human judgment as final control layer
As AI agents take on more of the execution workload, the role of human engineers and testers shifts towards oversight, governance and quality control.
“Developers shift from writing every line of code to supervising generation, validating architecture, and managing quality.”
Within the agent factory model, humans act as “the editors-in-chief of the factory,” reviewing outputs, identifying issues and ensuring alignment with intended outcomes.
This is particularly critical in financial services, where regulatory expectations around model validation, explainability and operational resilience remain high.
“Software development becomes a continuous, high-speed loop rather than a two-week sprint cycle.”
– McKinsey book
AI-generated outputs, including test artefacts, require rigorous human oversight to ensure they meet both technical and regulatory standards.
While the gains are significant,“10 times the speed at half the cost” and the potential for “20 times leverage”, the model depends heavily on strong engineering and testing foundations.
AI agents require precisely defined workflows, clear user stories, and “unambiguous” acceptance criteria, as well as detailed system context such as architecture diagrams, API contracts, data models, service boundaries, and nonfunctional expectations.
Without this structure, the effectiveness of AI-driven testing and development quickly breaks down. “You can’t ‘chat your way’ to production-grade software,” the authors wrote.
For banks, the message is clear. The AI-driven transformation of software development is as much about strengthening testing discipline, governance and resilience as it is about accelerating delivery.
NEXT 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
