What happened
At Fortune's AIQ Summit, Bank of America's Hari Gopalkrishnan said rushing to AI is a common mistake, noting the bank often picks simpler rules instead, and that its AI budget will double next year.
Why it matters
The caution comes with heavy spending, and the bank says it decides against AI 'plenty of times,' so buyers may be pushed to justify each use case on cost rather than novelty.
What to watch
BofA's go-slower stance hinges on whether control infrastructure improves enough to justify more autonomy, and readers should watch the AI expense budget doubling next year.
WHO IT HITSEnterprise IT and AI platform teams evaluating AI vendors now face a buyer — a large bank — that reviews every project against risk categories and routes work by task difficulty rather than defaulting to AI, which could shift how vendors pitch cost and control.
Summaries like this, in your inbox every morning.
The panel in New York put two regulated-data giants side by side, and both framed AI as something to be reached for carefully rather than reflexively. Gopalkrishnan described a process that starts with client needs and a process inventory, and he said the bank often decides against AI. That caution sits alongside scale: he said Erica has handled 3.6 billion transactions, and without it the bank would need 11,000 more people to answer the calls.
Moore's account of S&P Global leaned on an early bet — the 2018 purchase of Kensho — which she said has given the company an advantage and now sits at the center of the business. The July 6 split into Kensho Data & Platforms and Enterprise Solutions, and a chief client office created about two years ago, are the structural moves behind that. She also described a tier-one bank whose accuracy rose from about 60% to 98% after S&P helped bring content sets into production 'six times quicker,' which illustrates what S&P is selling beyond raw data: speed and auditability in regulated workflows.
The two diverged on framing, not on fundamentals. Gopalkrishnan said AI is not always the right answer, while Moore said the question is less about the right tool and more about reinvention. The practical test for buyers and vendors alike is likely to be whether the control and review infrastructure both described can keep pace with the spending they have already committed.
Pick your industry and the AI tools you use, and get news related to your work every day.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.
