Andy Markus Is one of TIME's 2026 Executives of the Year: Tech and Data - Time Magazine

Many companies are still tinkering with AI, but under Andy Markus, AT&T’s generative AI program is well beyond experimentation. The company has a 45 billion daily expenditure of AI tokens—the smallest units of data that a language model uses to process and generate text—and has more than 1,000 use cases in production. Those deployments are producing a five-fold in-year cash return on investment, says chief data and AI officer Markus.

Working at that scale required the telecommunications giant to reconsider from whom it buys its intelligence; around 40% of inference volume now runs on open-source models post-trained specifically for AT&T-specific work. The company’s proprietary AI Gateway routes requests between models according to quality, cost, and latency, allowing agents to be deployed across customer care, network troubleshooting, and software development. (According to the company, in some cases, this initiative has cut AI costs by as much as 90% without sacrificing accuracy.) Markus believes open-source models could eventually account for 70% or even 80% of AT&T’s total AI use. “We’ll go as far as the accuracy will allow us to go,” he says.