BNY builds tools to track AI token spending amid rising costs

BNY has developed internal tools to track its spending on AI tokens and optimize task routing across models, according to Chief Financial Officer Dermot McDonogh. The bank, which allocates $4 billion annually to technology, is running about 400 AI agents and facing rising costs despite falling per-token prices.

Tarun Mottlia

Editor, Lazyfounder

Published Oct 28, 20262 min read

BNY builds tools to track AI token spending amid rising costs — Image: Igor Omilaev on Unsplash via source

Tools to monitor AI token spending

BNY has built internal tools to track its spending on AI tokens and route tasks to the most suitable AI models, according to statements from Chief Financial Officer Dermot McDonogh. The tools were developed after the bank’s July comment to Fortune, where McDonogh said BNY did not track token use and instead focused on results.

The bank runs about 400 AI agents and allocates $4 billion annually to technology, with AI forming a portion of this budget. The tools aim to prevent overspending as AI costs rise across the industry.

Why AI costs are rising despite cheaper tokens

Rising AI costs are a growing concern for companies, including banks, even as per-token prices fall. The increase is driven by higher token consumption from AI agents, which now handle more tasks than before. McDonogh highlighted this trend as a key reason for BNY’s disciplined approach to tracking spending.

BNY sources AI models from major cloud providers and avoids dependence on a single platform, despite its CEO, Robin Vince, joining OpenAI’s board in July.

How this fits with broader AI adoption in finance

BNY’s focus on tracking AI spending reflects a broader challenge for institutions adopting large-scale AI. While cost management tools are not yet widespread, they may become essential as AI agents proliferate and token usage grows.

Earlier reporting from OpenAI’s product manager Ted Sanders outlined a framework for selecting GPT models based on performance, cost sensitivity, and technical needs. BNY’s approach aligns with this trend by emphasizing efficiency in model selection, though it applies the principle at an organizational level.

What this means

Lazyfounder analysis — our interpretation, not reported fact.

BNY’s move to track AI token spending is a signal for founders and operators in AI-driven industries. For banks and enterprises, AI costs are no longer just about model pricing—they’re about usage volume. Even as per-token prices drop, the sheer scale of AI agent deployment can lead to unexpected budget overruns.

This development underscores the importance of granular cost-tracking tools, especially for companies running hundreds of AI agents. Founders building AI-heavy products or services should consider whether their current infrastructure can monitor token consumption in real time. If BNY, a $4 billion tech spender, is treating this as a priority, smaller players may face even greater pressure to optimize spending before it spirals.

The trend also highlights a shift in how AI adoption is being managed at scale. Earlier this year, BNY’s CFO dismissed tracking token use, but rising costs have forced a change in strategy. This suggests that even for large institutions, AI spending is still being figured out—not just in terms of what works, but what’s financially sustainable.

Sources

The Next Web · 2026-10-27
BNY says it now tracks its AI token spending to avoid overpaying

This story is an original summary drafted with AI by Lazyfounder from the reporting listed above and checked by automated validation. Facts are attributed to their original publishers; sections marked as analysis are Lazyfounder's. Where a source is in another language, facts were machine-translated and quotations are reported, not reproduced. Read the original coverage via the links, and see our AI policy and corrections policy.

About the author