America’s biggest bank JPMorgan Chase is exploring a future where artificial intelligence agents will be able to negotiate, agree on terms and also execute payments with one another without direct human intervention, but the banking giant said that strict controls will be required to make sure that the technology tays within predefined limits. According to a report by CNBC, JPMorgan’s blockchain unit Kinexys and South Korea's Hyundai Card have already completed an initial proof-of-concept which tested how AI agents could handle financial transactions autonomously. The pilot also involved a simulated foreign-exchange transaction designed to help Hyundai Card maintain Euro liquidity in its blockchain-based deposit account.AI agents negotiating with each other is the next goal: JPMorganThe report further states that the next phase of the project will focus on testing what happens when multiple AI agents interact directly with a financial ecosystem, instead of simply carrying out predefined instructions. "What happens when agents are negotiating, agreeing directly with each other and then making payments? What does that actually look like?" said Wee Kee Toh, Global Head of Kinexys Insights at JPMorgan. Speaking to CNBC, Toh said the companies are targeting the first half of 2027 for a more advanced phase that explores autonomous agent-to-agent interactions. "That, I would say, is really the biggest part of what we want to get as an expansion of the phase one scope," he said.JPMorgan wants transactions to remain within strict guardrailsWhile the AI agents could eventually handle complex financial operations independently, JPMorgan feels that governance and oversight still remain the biggest challenge. As per Toh, developing the technology itself was not particularly difficult. The greater hurdle was figuring out accountability and managing the risks associated with allowing AI systems to move money without direct human supervision. In order to address these concerns, Kinexys has built a framework that confines AI agents to authorised parameters and automatically rejects transactions that exceed their permissions."You are going to draw a box around the agent. The agent will never be able to do anything out of the box," Toh said. For example, an AI agent authorised to execute transactions worth up to $5 million would be unable to complete any payment above that threshold.The approach is designed to keep humans out of the transaction process while ensuring agents cannot act beyond predetermined rules and controls.First blockchain-based autonomous treasury testTaking about the initial trials, JPMorgan described it as the first private blockchain-based autonomous bank-agent-to-client-agent test for corporate treasury management. The proof-of-concept included a Hyundai Card AI agent starting a foreign-exchange transaction over Kinexys' blockchain infrastructure without direct involvement. JPMorgan has still not announced a commercial deployment of the technology despite successfully completing the pilot.Partnership could expand across Hyundai groupAlong with this, the companies are also working on evaluating, if the collaboration could be extended to other businesses within Hyundai Motor Group. According to Toh, Hyundai Card's links to the wider group were one of the factors that made the company an attractive partner for the project. Hyundai Motor Group currently owns a majority stake in Hyundai Card through Hyundai Motor, Kia and Hyundai Commercial.Growing push into blockchain paymentsThe AI-agent initiative comes as both companies continue investing in blockchain-based financial infrastructure. Earlier this year, Hyundai Card said it completed a proof-of-concept involving a $20,000 USDT stablecoin transfer between Hyundai Motor's US and Mexican subsidiaries, with settlement and verification taking an average of seven minutes.Meanwhile, Kinexys said in August that it had processed more than $4 trillion in transactions since inception, averaging about $7 billion in daily transaction volume. JPMorgan has also been expanding its digital asset offerings, including the rollout of its dollar-denominated deposit token JPMD for institutional clients.For JPMorgan, the long-term vision is not simply using AI to automate payments, but enabling autonomous agents to negotiate and settle transactions with one another while operating inside tightly controlled boundaries designed to prevent unauthorised actions.
JPMorgan wants AI agents to handle transactions without human intervention
Full Article
Original Source
Read the full article at Timesofindia →KhanList aggregates and links to publicly available news content. We do not host full articles from third-party sources. Always verify important information with original sources.