# Personal AI agents identify themselves under 3% of the time in transactions

> Arkose Labs reports most AI agents act without clear identification when logging in, signing up or checking out for users.

Oossa · 2026-10-20 · https://oossa.com/en/personal-ai-agents-identify-themselves-under-3-of-the-time-in-transactions

Arkose Labs released its Q3 2026 State of Agentic AI Trust Report. The study finds that personal AI agents, including Meta’s Muse, reveal their identity in less than 3% of interactions that involve sensitive actions such as logging in, signing up or checking out. Over 65% of sites in banking, gaming, tech and travel reported agents performing these tasks on behalf of customers. The report notes that control mechanisms have not kept pace with this growth.

## The facts

- Agents identify themselves in under 3% of sensitive transactions.
- More than 65% of customer sites let agents log in, sign up, or checkout.

## Why it matters

Consumers may not realize an AI, not a human, is handling their login or payment, raising privacy and security concerns.

## Sources & references

1. [Personal AI Agents Including Meta’s Muse Identify Themselves Less Than 3% of Time While Handling Sensitive Transactions at Scale](http://www.businesswire.com/news/home/20261020106099/en/Personal-AI-Agents-Including-Meta%E2%80%99s-Muse-Identify-Themselves-Less-Than-3-of-Time-While-Handling-Sensitive-Transactions-at-Scale/?feedref=JjAwJuNHiystnCoBq_hl-WsGkFyUNMTSNrPsg57HLEqqcp-o_pnudlUwsb5apQ1S4gUE65BTfjH3-pSuqdv0gW3cb3F4oTIgUqCPafFkgu5AneicJQDNboWvodvG82E0) – Business Wire, 2026-10-20

Last updated: 2026-10-20
