If you are still debating whether to integrate AI into your core operational model, NatWest just dropped some numbers that might end the discussion.
We aren’t talking about small experiments anymore. As of 2025, the banking giant has moved past the “testing” phase into full-scale deployment, and the efficiency gains are turning heads. For business leaders looking for benchmarks, this is what meaningful AI adoption actually looks like.
The Developer Efficiency Jump
For any founder managing a tech team, this statistic is the headline: AI now produces over one-third of NatWest’s software code.
This isn’t just about autocomplete. Their 12,000 engineers are using AI for drafting, reviewing, and testing. In specific trials within their financial crime units, “agentic engineering”—where AI agents handle complex tasks autonomously—led to a tenfold increase in productivity.
The takeaway? If your development team isn’t leveraging AI, you are likely building slower and more expensively than the competition.
Reclaiming 70,000 Hours
In the retail division, the impact on the bottom line is tangible. By automating call summaries and complaint drafting, the bank saved over 70,000 hours of staff time.
This addresses a classic operational bottleneck: documentation. Instead of human staff spending hours transcribing calls or drafting standard responses, AI handles the heavy lifting. This allows the team to focus on resolving complex issues rather than getting bogged down in administration.
The “Agentic” Upgrade to Customer Service
The standard chatbot is evolving. NatWest’s digital assistant, Cora, has moved from simple scripts to generative AI, expanding its capabilities from 4 to 21 distinct customer journeys.
But the real news is the shift to agentic AI. Built on OpenAI models, the new system allows customers to ask natural language questions about spending patterns or recent transactions. It’s no longer about keywords; it’s about conversation. The roadmap even includes voice-to-voice capabilities that can detect tone and nuance—essential for sensitive interactions like fraud reporting.
More Face Time, Less Admin
In Wealth Management, the fear is often that technology will dilute the human touch. NatWest’s data suggests the opposite.
By using AI to summarize client meetings and manage records, relationship managers have freed up 30% more time. That is time redistributed from paperwork back to face-to-face client advising. AI isn’t replacing the relationship; it is removing the friction that prevents it.
The Verdict
NatWest has signaled a shift that applies to every industry: AI is no longer an “experimental adjunct.” It is a core component of the modern operating model. Whether it is writing 33% of your code or saving thousands of hours in admin, the metric for success is no longer “did we implement AI?” but “how much efficiency did we unlock?”
