Lloyds Banking Group is hiring 300 tech experts to work on agentic AI by September, weeks before CEO Charlie Nunn unveils a new multi-year strategic plan. The bank already uses Anthropic’s Claude and Google’s Gemini internally, AI delivered a £50 million boost last year, and the group expects £100 million this year. But the bank’s own head of data and AI says AI will reshape how the organisation is structured — and the CEO acknowledged in January that jobs will be cut.
🔍 THE BOTTOM LINE
The pattern is clear and it is not subtle: Lloyds is hiring the people who will build the systems that will replace the people they are not hiring. The 300 new recruits join a 1,000-strong AI team that includes retrained Lloyds staff. They will deploy existing LLMs like Claude and Gemini and build agentic systems on top of them. The hire-now-cut-later trajectory is the story — and it is the same trajectory every major bank in the Anglosphere is on, including New Zealand’s Big Four.
What Changed: From LLM Experiments to Agentic Deployment
Lloyds’ strategy has moved past proof-of-concept. The bank is now building agentic AI — autonomous models that can plan and execute multi-step tasks with minimal human oversight. Trystan Davies, Lloyds’ group head of data and AI science, said the new recruits will work on identifying and preventing scams and fraud, searching internal HR documents, and making online banking more personalised. Customers will be able to ask plain-language questions about their finances.
But Davies was also blunt about the implications: “AI will reshape how organisations are structured. It will change roles and how we work, and we are investing in training for colleagues through that transition.” That is corporate language for: some of those colleagues will not have roles to transition to.
The financial returns are real. Generative AI delivered £50 million to Lloyds’ balance sheet last year, and the group expects double that in 2026 from growing use of agentic models. When a 261-year-old bank says AI is a profit centre, not a cost centre, the strategic direction is locked in.
Context: The Banking Sector’s AI Pivot
Lloyds is not an outlier. Standard Chartered announced 7,000 job cuts earlier this year, partly attributed to AI integration. Its CEO Bill Winters later apologised for describing the move as “replacing, in some cases, lower-value human capital” — but the cuts proceeded.
The broader picture from KPMG’s latest financial services sentiment survey is one of overconfidence: 93% of UK bank executives believe they can keep operating during a significant AI outage, but only 47% have carried out a single test around AI disruption, and 26% have not conducted any. As KPMG’s Rob Smith put it: the industry’s optimism about resilience “could mean one of three things: firms have invested considerably in contingency planning; firms’ use of AI tools is relatively simplistic; or they don’t yet have a complete grasp of their exposure.”
NZ Angle: The Big Four Are on the Same Curve
New Zealand’s major banks — ANZ, BNZ, Westpac, and ASB — are subsidiaries of Australian banking groups that are themselves on the same AI adoption trajectory as Lloyds. ANZ announced an AI workforce transformation programme in 2025. The technology pressure points are identical: agentic systems for fraud detection, customer service automation, and internal document processing.
The “hire-now, cut-later” pattern is the one NZ workers should watch. Lloyds is hiring 300 specialists at high cost before the full automation impact hits its existing workforce. They need the engineers to build the systems first. Once the systems are built, the roles they automate are restructured. The timeline from “300 new hires” to “role restructure announcements” at Lloyds is likely 12-18 months — and NZ banks will follow on a similar lag.
For Kiwi finance workers, the skill shift is specific: from executing routine processes (compliance checks, loan screening, data entry) to managing, auditing, and prompting the AI systems that will execute those processes. The value moves from doing the task to knowing why the AI should act and how to verify its output.
The Other Side: AI Resilience Is a Real Risk
The KPMG data is the counter-narrative to the bullish hiring story. Banks are adopting agentic AI faster than they are testing for failure modes. An agentic system that fails due to prompt drift, data contamination, or model hallucination could cascade through customer-facing systems in ways that traditional IT outages do not — because the system does not crash, it just produces wrong answers confidently.
Only 47% of UK banks have tested for AI disruption. NZ banks have not publicly disclosed their AI resilience testing. If the financial sector’s track record on pre-crisis stress testing holds, the gap between adoption and preparedness is where the real systemic risk lives.
The Bigger Picture: From Cost Centre to Core Infrastructure
The Lloyds hiring drive confirms what the Gallup AI holdouts data already showed: workers who do not adopt AI face triple the layoff risk. But the Lloyds story adds a structural dimension — it is not just individual workers who need to adapt, it is the organisations themselves that are redesigning around AI as a core system, not a supplementary tool.
For NZ, the policy question is whether our banking regulator (the Reserve Bank) is tracking AI adoption in the financial sector with the same scrutiny it applies to capital adequacy. If 93% of bank executives believe they are resilient but only 47% have tested, the gap between confidence and evidence is a regulatory concern, not just an operational one.
❓ FAQ
Q: What is agentic AI and why is it different from what banks were doing before? A: Agentic AI systems can plan and execute multi-step tasks autonomously — booking an appointment, checking an account balance, and sending a confirmation without human prompts at each step. Banks previously used LLMs for single-task assistance (chatbots, document summarisation). Agentic systems replace workflows, not just tasks.
Q: Is Lloyds hiring 300 people to replace other jobs? A: In the short term, the 300 hires increase headcount. In the medium term, Lloyds’ own CEO has said AI will require “reducing some jobs in some areas.” The hires are the build phase; the cuts are the deployment phase.
Q: Are NZ banks doing the same thing? A: NZ’s Big Four are subsidiaries of Australian banks that are on the same AI adoption curve. ANZ, CBA, Westpac, and NAB have all announced AI transformation programmes. The specific pattern — hiring AI engineers before restructuring existing roles — is the standard playbook across the sector.
Q: What should a NZ finance worker do right now? A: Treat routine process execution as a declining asset. The durable skills are AI system management — prompt engineering, output auditing, edge-case handling, and understanding the regulatory implications of AI-generated decisions in financial services.
🔍 THE BOTTOM LINE
Lloyds is telling you exactly what is coming: hire the builders first, restructure the operators second. NZ banks are on the same trajectory. The window between now and when those agentic systems go live is the window for finance workers to shift from doing the work to managing the systems that will do the work. It is not a large window.
