The increasing use of artificial intelligence in financial systems is driving a shift towards machine-executed transactions, posing new challenges for control, oversight, and infrastructure, according to Microsoft and Chainalysis executives. Bill Borden, Microsoft's corporate vice president of worldwide financial services, warned that legacy systems will soon face significant pressure as transaction demands become more complex, with the tipping point occurring when latency, scale, and complexity start to impact competitiveness. This will force firms to reevaluate their system architecture, he said at an Alchemy-hosted event in New York City. While automation has long been present in finance, Borden noted that the focus is now shifting from capability to trust.
The question is no longer about technology's ability to automate tasks like executing hedging strategies, but rather about trust, audit, and control. To facilitate this transition, Microsoft is developing tools to manage machine identities and permissions, as well as track AI agent actions. In regulated environments, firms must demonstrate what controlled a system and whether it adhered to policy when decisions are made without direct human input, Borden said.
Jonathan Levin, co-founder and CEO of Chainalysis, pointed to the crypto sector as a working model of automated finance, where blockchain networks process large volumes of transactions through smart contracts and software-driven wallets, creating an environment similar to agent-based systems. The crypto sector has been preparing for this moment long before other parts of the financial services industry, Levin said.
This experience extends to risk management, with Levin citing efforts to track illicit funds across thousands of wallets as an example of the kind of monitoring needed in systems where transactions occur at scale without direct human input. Looking ahead, both executives anticipate a mix of systems to coexist.
Levin predicted that the majority of commerce in 10 years will be settled on public infrastructure, while Borden envisioned a more integrated approach linking public blockchains, private networks, and existing rails. Traditional rails will continue to exist, Borden said, with software serving as the connecting layer.