AI agents to become most powerful identities this year

Token Security, an artificial intelligence security company, said AI agents will become the most privileged identities inside large organizations in 2026, creating new security and compliance risks as companies move these systems into live business operations, according to Security Brief United States.

The company expects enterprises to deploy AI agents at scale across finance, software development, HR, and customer operations, where they will execute actions, access sensitive data, and trigger workflows. It predicts agents will outnumber human users and receive broader permissions, requiring changes to identity management and regulatory frameworks that currently assume humans are at the center of most processes.

Itamar Apelblat, CEO and co-founder of Token Security, said organizations will need clearer controls and accountability for agent activity, while the firm's CTO and co-founder Ido Shlomo warned that continued use of static API keys and long-lived credentials will weaken agent security. The firm also expects each employee to rely on multiple agents, shifting identity risk from user accounts to the systems acting on their behalf.

Token Security further forecasts new security incidents caused by mismatches between human access rights and agent permissions, as well as misconfigurations introduced by AI coding tools, with agent identity failures driving future breaches.