Financial services firms are moving rapidly to adopt artificial intelligence but remain poorly prepared for the workforce changes it could bring, according to a PwC survey of more than 1,000 senior executives.
The survey found that 42 per cent of financial services leaders have carried out company-wide modelling of how AI could change the amount of labour they need, while almost eight in 10 expect their workforces to shrink by at least 20 per cent over the next five years.
PwC said the figures point to a gap between calculating how many jobs could disappear and planning what the workforce will actually need to look like as AI becomes more widely used.
Among companies that have begun workforce modelling, only half have examined how AI-driven changes to processes and workflows could affect their staffing needs.
The findings come from PwC’s 2026 Financial Services Workforce AI Survey, which examined how US financial services firms are preparing for AI and its impact on hiring, skills, pay and leadership.
The survey found that financial firms are under pressure to accelerate their adoption of AI, with 90 per cent of executives saying companies need to become more comfortable moving quickly.
At the same time, 77 per cent said their own organisation was not moving quickly enough to keep pace with AI innovation, despite 70 per cent saying their company was already moving faster to remain competitive.
Regulatory and compliance requirements are one obstacle, but employee concerns are also slowing progress.
Some 44 per cent of executives said employees were worried about job security or changes to their roles, while 43 per cent said workers used AI only when required rather than taking the initiative.
Another 40 per cent said employees felt overwhelmed by the pace of AI-driven change, while 34 per cent identified change fatigue as a major barrier to expanding AI use across their workforce.
At the same time, firms are placing a growing financial value on employees who can work effectively with AI.
Some 91 per cent of executives said their companies were increasing pay for employees with AI skills, while 58 per cent planned to link compensation directly to productivity gains achieved through AI.
A further 86 per cent said AI skills training was more valuable than an MBA for many new hires.
To obtain those skills, 62 per cent of firms plan to hire employees with specific AI expertise during the coming year, while 61 per cent intend to retrain or develop existing staff.
Another 57 per cent plan to work with outside suppliers or service providers to obtain the capabilities they need.
PwC’s wider research suggests that AI will not affect all jobs in the same way.
Some roles are becoming more valuable because AI increases the importance of human judgement, critical thinking, team-building and creative problem-solving.
PwC describes these as “professionalised” jobs, saying they are growing twice as quickly as jobs where AI makes specialist tasks easier for non-experts to perform, while wages in the former group are rising 42 per cent faster.
Despite the investment, the financial returns from AI remain uncertain.
Almost half of executives said improving productivity was a primary workforce objective, while 48 per cent were focused on reducing time spent on routine work and 46 per cent on incorporating AI into everyday workflows.
Technology and software engineering, risk management and operations were identified as the areas where executives expected the biggest productivity improvements.
However, 77 per cent said most of their AI investments were not producing measurable returns on investment.
PwC said firms should therefore establish clear performance benchmarks and financial targets before investing, rather than adopting AI simply because competitors are doing so.
Data is another major obstacle, with 41 per cent of executives identifying fragmented or poor-quality data as the biggest barrier to expanding AI across their workforce.
The survey also highlighted growing concerns around AI governance.
Almost nine in 10 executives said their firms had clear ownership and accountability for decisions involving AI agents, but there was little agreement over who should ultimately be responsible when an AI system causes significant harm.
The CEO and board were identified by 27 per cent, technology leaders by 16 per cent, risk and compliance leaders by 15 per cent and business unit leaders by 12 per cent.
PwC also found widespread use of so-called “shadow AI”, where employees use AI tools that have not been approved or centrally governed by their employer.
Ninety per cent of executives said employees using AI products outside centrally governed systems created regulatory risk.
PwC said firms will need clearer responsibility for AI risks, formal processes for introducing new systems and controls over which AI tools employees can access.
The survey covered 1,004 director-level and above executives at US financial services companies with at least $500 million in revenue, surveyed between May 12 and May 22, 2026.
Respondents were evenly divided between asset and wealth management, banking and capital markets, insurance and private equity.
