DNB cuts 400 jobs due to AI; HSBC plans UK layoffs

DNB, Norway’s largest bank, announced on October 6 the elimination of around 400 full-time positions, nearly 4% of its workforce, primarily within its Technology and Services division. The following day, the Financial Times revealed that HSBC plans to cut about half of its management and specialist roles in its wealth management business in the UK, with reductions of up to 70% among financial advisors. Both banks attribute these changes to artificial intelligence.

In the same week, Evident published its annual AI maturity index for banks, concluding something seemingly contradictory: institutions investing heavily in AI continue to expand their workforce.

Both interpretations are correct. The mistake lies in comparing them by total job numbers.

What DNB and HSBC Said

DNB's statement is explicit about the mechanism. The bank states that it has adopted agent-based AI in various areas, replacing tasks previously performed manually, with "substantial efficiency gains" in customer data management, KYC checks, technology development, and programming. "We are already seeing significant gains, and therefore we are adapting our organization to a new reality," said CEO Kjerstin Braathen. The reduction is expected to be completed by Q4 2026.

At HSBC, the figures come from sources interviewed by the FT and have not been confirmed by the bank, which does not disclose how many people work in the unit. The consultation process with employees is already underway, and those affected are expected to leave by the end of October. This plan represents a reversal: two years ago, the bank launched a hiring campaign specifically to expand its wealth management and private banking business in the UK.

The Counterargument, from Those Who Support It

Alexandra Mousavizadeh, co-founder and co-CEO of Evident, is the clearest voice of the opposing thesis. "The banks that are most dedicated to and spend the most on AI are the ones that are hiring the most," she told American Banker. "We see absolutely no impact on the headcount. Literally zero." The index places JPMorgan at the top for the fifth year in a row, followed by Capital One and Royal Bank of Canada, and highlights JPMorgan, Commonwealth Bank of Australia, and RBC as the largest employers.

This argument deserves to be taken seriously, with one caveat: the data comes from an index that Evident produces and markets as a sector benchmark, without cross-referencing independent employment statistics.

The cases of DNB and HSBC also have weaknesses as proof of substitution by AI. Analysts following DNB's announcement recall that the bank is already executing a cost-reduction program for the period from 2025 to 2027, and a bank pressured by costs has an incentive to present cuts as modernization. DNB did not detail how the 400 positions are divided by role, which hinders measuring how much of the cut correlates to tasks actually taken over by agents. HSBC, for its part, has dropped from eighth to eleventh place in Evident's index this year, making it difficult to present it as an example of a bank cutting because it has already mastered the technology.

The Right Focus is on Function, Not Total

The distinction that the debate often overlooks is in the type of work affected. DNB's cuts affect processing tasks: verification, data maintenance, routine coding. The hires that Evident measures are in AI engineering, governance, and deployment. A bank may cut 400 people in one area while hiring 400 in another, and a zero-sum balance obscures who lost jobs and who gained.

HSBC's case adds a new layer. Financial advisory for high-net-worth clients was seen as relationship work, protected from automation. If the 70% reduction is confirmed, it will be the first major layoff by a global bank in a frontline function, rather than a backend one.

Where the Effect Appears Outside Norway and the UK

The tasks DNB mentions, such as KYC and data maintenance, are the same that large European banks have outsourced over the past two decades to shared service centers in India and Poland. HSBC itself maintains global service centers in both countries. When a bank automates KYC in-house, the first volume to shrink tends to be outsourced or relocated to these hubs, not at headquarters, because that’s where the work has been standardized enough to be automated.

For consulting firms and BPO providers serving banks, the business question changes. The operational KYC contract charged by position loses its meaning when the client measures productivity by AI agent. Evident may be correct about the total number of jobs at leading banks, yet the jobs disappearing may be in Hyderabad or Krakow, outside the payroll that the index measures.