
N3XT's MCP protocol lets AI agents query balances and draft payments inside the corporate firewall. The FCA and OCC are watching the write-access feature.
N3XT, a blockchain-based narrow bank, has launched a Model Context Protocol that lets artificial intelligence agents read and write live banking data without leaving the corporate network.
The protocol means an AI assistant can query account balances and draft payments. It can also trigger escalation workflows in response to a natural-language instruction from a treasury operator. The data stays inside the company's firewall. Existing compliance rules, including maker-checker controls, still apply.
Aurélien Bonnel, N3XT's chief technology officer, said the company is standardizing how AI agents access live bank data. "We are enabling corporate clients to work within their own environment with data they can trust," he said.
The write capability is the part that draws regulatory attention. In the UK, the FCA has flagged automated decision-making in financial services as an area of focus. In the US, the OCC's guidance on third-party risk applies to any corporate client deploying an AI agent with payment-drafting authority. N3XT's emphasis on firewall containment and existing permission structures is a direct response, though clients will need to verify that the maker-checker workflow holds up under agentic automation.
N3XT sits in a narrower competitive set than a conventional treasury management system. Its direct rivals are programmable payment infrastructure providers and stablecoin-adjacent corporate banking platforms. The durability of its advantage depends on whether enterprise treasury clients prefer keeping AI workloads inside their own environment, given the sensitivity of real-time payment data.
The next markers to watch are named enterprise deployments and independent security audits of the MCP layer. Whether N3XT pursues a US bank charter will also shape its trajectory.
Drafted by a large language model from the source reporting linked above, then screened by automated publishing checks. It is not read by a journalist before publication. Some articles cite our Alpha Score. Verify prices and figures against the original source. Educational coverage, not personalized advice.