Major banks shift AI focus from usage metrics to financial returns

Bank of America has disclosed specific financial returns from its artificial intelligence initiatives, marking a significant shift in how major financial institutions report on technology investment. Chief Executive Officer Brian Moynihan stated that the bank currently operates between 130 and 140 AI projects. These initiatives have incurred a cost of $400 million while generating benefits of $800 million. Although Moynihan did not specify the exact time period covered by these figures, the disclosure represents a move away from previous reporting standards that focused primarily on usage counts and hours saved. The bank intends to double its AI expense budget in the coming year, reflecting a continued commitment to expanding its technological capabilities.

This strategic emphasis on measurable financial outcomes aligns with broader industry trends where leaders are increasingly quantifying the value of AI in monetary terms. Bank of America allocates $4 billion annually to new technology initiatives, including AI, from a total technology budget of $13.5 billion. The tangible benefits of this investment are evident in customer service operations. The bank’s virtual assistant, Erica, had 24.6 million active users in the second quarter. Chief Technology and Information Officer Hari Gopalkrishnan noted that without this tool, the bank would require an additional 11,000 staff members to handle call volumes. Furthermore, EricaAssist, a generative AI tool used by more than 18,000 customer service representatives, has reduced average call times by nearly one minute. In July, Chief Financial Officer Alastair Borthwick reported that the bank had approved approximately 300 AI use cases and deployed more than 100 across its operations.

European banks are similarly tying their AI strategies to concrete financial targets. Société Générale expects artificial intelligence to contribute up to $669 million to its planned savings by 2029. This figure forms part of a broader cost plan aiming for gross savings of approximately $2.1 billion between 2026 and 2029. The French bank is targeting a cost-to-income ratio below 55% by 2029 and has signed a strategic agreement with Anthropic to accelerate AI adoption. Meanwhile, Commerzbank in Germany has set out its own AI goals as part of its Momentum 2030 strategy update. The lender plans to invest about $669 million in AI from 2026 to 2030. Chief Executive Officer Bettina Orlopp described AI as a key driver of the bank’s transformation, noting that the investment should save approximately $390 million annually by 2030. This is expected to free up about 10% of staff capacity for redeployment, allowing sales teams to focus more on client advice. The bank is targeting a cost-to-income ratio of 43% by 2030.

Industry data suggests that while returns are becoming more common, full payback remains a longer-term goal. A report by PYMNTS Intelligence indicates that between 90% and 100% of large U.S. enterprises report some return on AI spending. However, only 5% to 10% report a full payback. The survey, which included 60 senior technology executives at companies with at least $1 billion in revenue, found that at least half of firms in every industry group expect full payback to take five to six years. Chief Financial Officers are increasingly measuring AI returns through improved margins and lower operating costs. The share of CFOs tracking higher revenue per customer has risen significantly, jumping to 56.7% from 18.3% in July 2025. This shift underscores a growing emphasis on financial performance metrics in evaluating the success of artificial intelligence initiatives across the corporate sector.