Transmit Security has introduced Agentic Trust, a service intended to help banks and merchants identify AI agents and control what they can do. The company has also appointed Matt Thompson as chief executive, replacing co-founder Mickey Boodaei, who is to remain involved in product development.
Watch Desk analysis
What happened
The announcement, reported by Biometric Update on 19 October, marks a move beyond identity systems built chiefly to distinguish people from malicious automated traffic. Transmit Security says Agentic Trust is designed to detect agent-driven activity, classify the agent and its apparent actions, then let organisations investigate and apply policies.
Those policies could allow or restrict activity, require customer approval for higher-risk actions, or block transactions. The company also says permissions can be scoped and time-limited. That matters when an authorised assistant may be able to make a purchase but should not be able to transfer money or change account details.
Why it matters
AI agents acting through valid customer credentials complicate the old security question of whether traffic comes from a person or a bot. The harder question is what the agent is authorised to do on someone’s behalf. Transmit Security is pitching detection and governance as a way to address that gap, including when an agent does not identify itself.
The commercial stakes are clear for financial institutions and merchants, but the announcement leaves important evidence outstanding: it gives no independent detection-accuracy results, customer deployment figures or detailed technical explanation of how the system verifies an agent’s relationship to a customer.
Our read
This is a concrete attempt to treat AI agents as a new kind of actor in digital identity, rather than squeezing them into the human-versus-bot box. The sequence of detect, classify and govern is sensible; whether it works reliably is the part that still needs showing. Buyers should look for evidence that the system can distinguish legitimate delegated activity from fraud, not just a tidy set of policy controls.
What to watch
- Whether Transmit Security publishes detection results and explains how it verifies delegated authority.
- Whether banks or merchants deploy Agentic Trust and disclose what it can control.
- How the company’s approach compares with competing tools focused on agent identity, authorisation or transaction risk.
Discussion spark: Should businesses first verify an AI agent’s identity, or focus on enforcing what it is permitted to do even when its identity is unclear?
Sources and evidence
- Source update (19 October 2026, 13:03 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.