At the United Nations Headquarters in New York, BeInCrypto led a regulatory panel during the Future of Money summit on AI's impact on finance. The panel featured Dottie Romo, chief risk and control officer at the US Internal Revenue Service; Dino Cataldo Dell'Accio of the UN Joint Staff Pension Fund; and Julius Moye, manager at Mastercard's Financial Crime Solutions. The discussion centered on whether current oversight models can keep pace with AI-driven financial activity.
Romo: periodic reports can't catch AI-speed fraud
Romo said regulators still depend heavily on periodic reports to spot fraud and control failures, a process she described as too slow for autonomous finance. "They're making millions of decisions in minutes," she said, referring to AI systems. Romo argued that some form of automated supervision will likely be necessary. "We're not going to be able to do that in a fast enough pace," she said, advocating for more real-time monitoring while keeping humans involved in important decisions.
Moye points to Knight Capital and Terra/Luna as cautionary tales
Moye cited the 2012 Knight Capital trading disaster and the collapse of Terra/Luna as warnings about automated systems running without sufficient safeguards. He proposed a model where machines detect unusual behavior and automatically contain the problem, with serious incidents escalated to humans. "It's really using AI and machines to apply the tourniquet and stop the bleeding and then have humans come in to do the surgery," Moye said.
Dell'Accio: accountability still traces to three questions
Dino Cataldo Dell'Accio argued that automation cannot erase responsibility. He said accountability should trace back to three questions: "Who developed the code? Who implemented the code and who oversees the code?" The framework, he suggested, keeps human roles central even as machines take on more decision-making.
BeInCrypto research shows $119,947 across 6.4 million x402 transactions
The panel came as BeInCrypto published its State of AI Agent Payments 2026 research, which tracked 6.4 million x402 payment transactions carrying $119,947 across Base and Solana between July 23 and August 26. Of those transfers, 90.8% were worth less than one cent. The research also noted that one AI agent recorded nearly 200 million settlement transactions since launch. The data highlights the high-volume, low-value nature of agent-driven payments.
The panel left open exactly how automated supervision would be implemented or which regulators would take the lead. No timeline was announced. The next concrete step would likely involve pilot programs or further research from the participating organizations.
