AT&T uses agentic AI to reinvent business processes from the ground up

The business move

AT&T is shifting from using AI to simply boost employee productivity to embedding autonomous AI agents directly into its core business workflows. The telecom giant now runs around 500 agentic AI workflows in production and plans to triple that by next year. This represents a move from isolated AI tools to system-wide process reengineering powered by autonomous agents. The company processes roughly 45 billion AI tokens daily, showing significant scale in its AI operations.

Why it matters

Turning AI into self-directed agents that redesign how work gets done changes the game for large enterprises. AT&T is betting on this approach to cut costs, speed operations, and unleash new efficiencies without relying solely on human supervision. This means AI is moving beyond a productivity booster to a foundational layer that shapes company processes from the ground up. For enterprises, this raises the bar for AI deployments: success is no longer just about getting AI to assist people but getting AI to independently manage business workflows.

Who gains and who gets squeezed

AT&T’s aggressive push forces competitors to rethink their AI strategies and invest in autonomous agents or risk falling behind. Vendors that provide agentic AI tools and platforms stand to gain from increased enterprise demand. Meanwhile, companies relying on manual or semi-automated processes will face pressure to accelerate automation or lose efficiency. Internally, roles focused on routine workflow tasks could diminish as AI agents take over, shifting workforce needs toward oversight and strategy rather than execution.

What to watch next

Watch to see if AT&T can manage scaling agentic AI workflows without imposing new operational risks such as automation errors or increased complexity. How the company balances AI autonomy with human control will be key. Also monitor what types of workflows AT&T targets next—whether they remain in back-office functions or expand to customer-facing and network operations. Others in telecom and beyond will look for proof points before committing to similar autonomous AI overhauls.

AI Quick Briefs Editorial Desk