Sumsub and Sumvin Partner to Close the Identity Verification Gap for AI Agents

The next frontier in AI security isn't the model — it's proving who is behind it. As AI agents begin buying things, opening accounts and managing money on behalf of people, Singapore's financial sector is waking up to a question the industry hasn't fully answered: when an agent transacts, who exactly is accountable? A new integration from identity verification firm Sumsub and agentic credential startup Sumvin points at how the industry plans to answer that, and it has direct implications for the banks and fintechs building agentic services here.

Sumsub and Sumvin have partnered so that AI agents can carry a portable, encrypted credential linking them to a verified human. Sumvin's agentic credential will be powered by Sumsub's AI Agent Verification technology — part of Sumsub's Know Your Agent framework — alongside its Reusable Identity tools. Once a person completes know-your-customer checks, their verified identity can travel across platforms, letting an agent demonstrate both who it represents and that it has permission to act, without forcing the user to redo onboarding at every new service. The developer offering is expected to become available globally this month.

The integration targets a specific gap: agentic commerce — purchases and financial activity executed by AI agents on a user's behalf — has grown faster than the trust infrastructure around it. Today, a merchant or financial institution dealing with an agent has limited ways to confirm the identity, authority and intent behind that agent. Sumvin's CEO Simon Jones frames the credential as "the missing primitive" for agentic commerce, while Sumsub's Thomas Taraniuk describes the work as extending KYC-grade trust "into the agent layer" — essentially treating the agent as an extension of a verified person rather than a separate, anonymous actor.

For Singapore, this lands at a moment when local banks are pushing agentic AI into real customer workflows — from wealth client onboarding to AI assistants that execute banking tasks — and when regulators are scrutinising how accountability works when machines act. The Monetary Authority of Singapore has repeatedly stressed that financial institutions remain responsible for outcomes of AI systems they deploy, which makes verifiable agent identity a practical compliance building block, not just a product feature. It also hints at the direction of travel: fewer repeated onboarding loops, and a portable identity layer that agents carry between institutions.

Why it matters for Singapore: Singapore's fintech ambitions depend on trust, and agentic AI is about to test that trust at scale. If agents are to handle banking, payments and investment tasks for consumers here, the ecosystem needs a standard way to bind each agent to a verified human — which is exactly the problem Sumsub and Sumvin are attacking. The shift from re-verifying people at every platform to carrying one verified credential across services would reshape onboarding economics for local fintechs, and puts Singapore's banks in a position to adopt the pattern early as they scale agentic services.