Most consumers are not ready to let AI agents shop unsupervised

Okta's 2026 Customer Identity Trends Report indicates that consumer trust in AI agents that make purchases, book services and access personal information depends on agent capability, the consequences of mistakes, and the ability to approve or reverse actions. The report says businesses must meet new requirements for agent verification and permission management. The article reports that only 37% of respondents are ready for this, but the sentence is truncated, so the population and wording behind that figure cannot be confirmed. No vulnerability or specific control guidance is given.

For security analysts · Written by AI from 1 source · Wrong?