What happened
On Oct 19, Transmit Security announced Agentic Trust, an offering (unveiled alongside a new CEO) that detects AI agent traffic across online channels — including agents that don't identify themselves — classifies them (AI assistants vs custom automation frameworks), and lets organizations assess activity, apply governance policies to permit/restrict/block actions, require customer approval for higher-risk actions, and scope permissions by time and purpose. It targets financial institutions and merchants where agents open accounts, transfer funds and complete purchases with valid customer credentials.
Why it matters
This addresses the CIAM/fraud-adjacent gap where AI agents act on a customer's behalf using legitimate credentials, evading existing human/anti-bot controls. It gives FIs and merchants a way to distinguish authorized AI activity from risky automation and enforce granular, approval-gated control rather than allow-all or block-all.
Applicability
Financial institutions and merchants with consumer-facing AI assistants should evaluate Agentic Trust to detect and govern agent-driven transactions, especially for account opening, payment transfer and purchase completion where customer approval on high-risk steps is required.