Socure, the identity verification and fraud company, launched Socure Agent Intelligence on 27 October 2026, a suite that detects AI agents in browser sessions, scores whether to trust them and binds each agent to the verified person or business that authorised it, according to the announcement on BusinessWire. Fintechbits reported on 16 October 2026 that Socure’s Q3 release gave $42.7 million in new bookings and more than 6,000 customers but no ARR figure, so the company is launching new products without updating the revenue metric it reported in Q2.

How Socure Agent Intelligence works

Agent Detection, available inside Socure’s Digital Intelligence product, identifies agents driving browser sessions whether or not they declare themselves, and returns a confidence score and reason codes. Socure says named consumer agents were identified with 99.99% accuracy across 700 behavioural variables in controlled lab testing, a figure from lab testing rather than live traffic.

The release leans on demand figures from Accenture’s 2026 Consumer Pulse report, which it says found 85% of consumers open to collaborating with an AI agent and 74% open to having one carry out specific tasks. Socure Agent Intelligence is aimed at the businesses those agents will try to transact with.

Agent TrustScore combines identity, authorisation and risk data, and is open to beta customers. Agent Credential is a W3C Verifiable Credential that agent operators issue through SocurePass to bind an agent to a verified principal, and Socure says any site can check it without calling Socure. Socure Agent Intelligence runs on the same identity graph Socure says handles more than 11 billion identity decisions a year.

“Block them all and you block your next customer,” said Johnny Ayers, founder and CEO of Socure. Other identity vendors are going after the same problem, as Incode’s TrustGraph network showed. Socure’s advantage is the identity graph it already runs. The weakness is that two of the three layers are still in beta or on the roadmap.