The U.S. Treasury Secretary, Scott Bessent, has brushed aside worries that the current surge in artificial intelligence could lead to a collapse similar to the dotcom crash. During the J.P. Morgan / Robin Hood Investors Conference in New York City, Bessent assured attendees that the AI sector isn’t facing a systemic downfall, despite some investors’ apprehensions, as reported by the Financial Times. Nevertheless, major financial institutions are taking measures to safeguard their customers from possible breaches linked to AI agents.
Bessent remarked, “I’m good at figuring out when there are problems in systems,” according to three listeners at the conference. He expressed confidence, stating, “I don’t see a systemic problem.”
He drew a contrast between the current AI landscape and the internet startup boom that preceded the dotcom bubble’s burst in the early 2000s, referencing failed ventures like Pets.com that were part of that era.
Bessent asserted emphatically, “This isn’t Pets.com. The smartest people in the world — who have 99 percent of their net worth in these companies and have built the most profitable companies in the world — they believe in it.”
In a related interview with Axios, he noted that today’s tech giants—like Microsoft, Google, and Meta—offer a different equation compared to the startups of the dotcom era. He highlighted that these large corporations are “the most cash-rich companies, probably in the history of the world,” and are actively investing from their own resources.
The Treasury Department did not reply to inquiries for additional comments.
In response to unauthorized access incidents involving AI agents, several prominent U.S. banks are implementing restrictions. Axios noted that banks like JPMorgan Chase urged clients to maintain the confidentiality of their login details, promoting safe data-sharing practices via authorized channels instead.
Wells Fargo confirmed their proactive approach to blocking unauthorized AI access, while Morgan Stanley reported no issues from AI agents on their platforms so far. As for Bank of America and Citi, no immediate comments were received, although it is known that they grant AI agents read-only access to customer data.
