Morgan Stanley doubles estimate: AI could eliminate 400,000 jobs in European banks by 2030

Analysts at Morgan Stanley published an updated research report on 28 May, raising the number of jobs at risk in the European banking sector due to artificial intelligence to 400,000 by 2030, up from the previous estimate of 200,000. Bloomberg reported on the document, which represents a revision of the initial forecast made in January, when the bank projected cuts equivalent to 10% of the workforce in the sector. The new projection indicates 20%, based on a central premise of a 30% productivity gain attributable to AI. The analysts themselves acknowledge that this conversion ratio has not been demonstrably proven at scale in the European banking sector.

The analysis covers 35 banks. The most exposed positions are in operational back-office, compliance, identity verification (KYC), anti-money laundering (AML) monitoring, and middle-office risk management. According to Morgan Stanley, the cuts would largely occur through natural attrition, early retirements, and negotiated exits over five years, rather than a single mass layoff event. The expected outcome is a structural reshaping of the workforce: data engineers, AI platform specialists, and model risk professionals enter, while back-office operation processors and traditional compliance officers exit.

From forecast to execution

UBS, absorbing over 30,000 staff inherited from the acquisition of Credit Suisse in 2023, is undergoing another round of cuts in Switzerland, which the bank frames within a US$10 billion cost-saving programme. ABN Amro, the Dutch state-controlled bank, has announced reductions in its credit and compliance divisions where AI systems have taken over some analyses. HSBC, headquartered in London, even considered eliminating 20,000 positions over a multi-year restructuring, as reported by Bloomberg in March.

Standard Chartered revealed their plans more explicitly than its peers. At an investor meeting in Hong Kong on 19 May, CEO Bill Winters stated that the bank will cut more than 15% of corporate support positions by 2030, equating to about 7,800 out of 52,000 back-office roles. "This is not about cutting costs; it is about replacing, in some cases, low-value human capital with financial and investment capital," Winters said. This statement encapsulates a stance that other European CEOs share in private but hesitate to vocalise publicly.

What weakens the argument

The premise of a 30% gain in productivity convertible into headcount cuts still lacks robust empirical evidence. Morgan Stanley itself acknowledges this in the note. A Gartner study published in May 2026 concluded that organisations that have made personnel reductions driven by AI are generally falling short of expected ROI: savings on payroll create budgets, but do not automatically generate business returns as end-to-end processes are rarely redesigned alongside the replacement of people. Without process redesign, adding AI over existing workflows accumulates licensing costs without achieving lasting headcount reductions.

The second factor of resistance is regulatory. The European Central Bank and national supervisors in Germany and France are assessing whether the automation of compliance and AML functions by AI introduces governance risks that offset cost gains. None of the 35 banks in Morgan Stanley’s scope had published a formal regulatory approval framework for the large-scale replacement of human compliance auditors with AI systems by the time of this report.

Where the impact is distributed

For Germany, Morgan Stanley's revision places additional pressure on Deutsche Bank and Commerzbank, which are pursuing efficiency targets via AI but have not published comparable headcount targets to Standard Chartered or HSBC. The absence of explicit disclosure does not imply a lack of movement: both banks have advanced their automation programmes in compliance and credit operations to advanced piloting stages.

In the Netherlands, ABN Amro and ING maintain significant portions of their back-office operations outsourced to Poland and the Czech Republic. Cuts in back-office functions at the parent banks are likely to reverberate in these shared service centres, where the same tasks are executed at scale and lower cost by local teams.

In India, where European banks maintain Global Capability Centres for compliance, risk, and financial operations in cities like Pune and Mumbai, the projection of 400,000 at-risk positions has a direct reading. The functions most exposed to replacement by AI are precisely those that European banks have historically transferred to India in search of scale and cost. A contraction in these areas affects not only the offices in Frankfurt, Amsterdam, and London but also the offshore operations centres that support them.

The question that Morgan Stanley does not resolve in the note is the one that matters most to European boards: will banks that cut 20% of their workforce by 2030 genuinely be more competitive or merely leaner? Standard Chartered, with Winters' statement, has bet that analysts and investors will be able to distinguish between the two. The next two or three quarters of results will reveal whether this bet is sound.