Forward-looking leadership: How this telecom giant puts AI into action

TOGETHER WITH

Future-forward: PwC helps businesses of all sizes expand their potential, and as a result, their opportunities. See how you can team up with their team of experts to turn your business challenge into an array of solutions.

What’s possible with applied AI? You could probably write a novel (or 10) about the business potential. But the thing is, we’re past the point of discussing potential. Now’s the time for strategy, applied learnings, and actionable insights.

AT&T’s leadership felt the same way. So when they spotted an opportunity to implement a tool to turn industry chatter into actual business action, they took it. And it’s not just about AI being the latest conversation starter for AT&T. It’s about their forward-thinking leadership spotting any variety of opportunities proactively, guiding the organization’s prioritization, and creating measurable business impact while others are still experimenting.

PwC sat down with AT&T at Mobile World Congress to understand their story and success not just with AI, but also with staying ahead of the curve with the challenges, trends, and in turn, opportunities that come their way.

*Quotes have been edited for length and clarity.

Making it happen: Measurable business impact

Implementation and education begin with an organization’s leaders. At Mobile World Congress, AT&T shared the scale, metrics, and strategy surrounding their AI program. Andy Markus, Chief Data and AI Officer at AT&T, led the discussion with PwC.

Markus noted that AT&T’s executive leadership was the driving force behind their successful AI introduction. For AT&T, focus beats volume, and governance is the key that enables scale. AI must be embedded into workflows, not layered onto legacy processes. Like with anything, you can’t just throw AI at the wall to see what sticks. Adoption matters as much as the technology in question.

AT&T’s leadership kept all this in mind to scale how AI is used across the entire organization. “They understand what [AI] should be doing for the company,” Markus explained. “We see a lot of companies fail across industries when the risk organizations are brought in late, and it creates chaos. Even some good ideas can’t see the light of day because they’re not the right risk profile for that organization.

“When we look at all the use cases across the business, for the ones that seem to resonate, we do a formal business case and look at the return on investment. And then we focus on the ones that have the highest ROI and prioritize that one to end, and make sure those are the ones that are executed first or the ones that drive value to our customers.”

Andy Markus, chief data and AI officer at AT&T, talks about business partnership ROI.

And as noted, because of all this groundwork, AT&T can focus on actually scaling AI, not just experimenting with it like many other organizations. Focused prioritization has allowed their company to start with intelligent governance, implement workflow redesign and adoption without disruption, and witness measurable business impact as a result.

This kind of understanding also allows for coordination between teams like finance and supply chain, in AT&T’s case, to further help confirm the organization is moving in lockstep.

Not every experiment will work, but with the right safety net, trying becomes more valuable than failing is costly.

Leadership-driven outcomes: The formula

While some companies are only beginning to dive into AI right now, AT&T has formulated a full roadmap. They’re not in the experimentation phase anymore. They’re already focused on scaling. And if there’s one thing AT&T is abundantly clear about, it’s that they’re not a pilot company.

Their leadership focuses on the things that truly drive value for AT&T and their customers. So when AI started becoming more and more of a potential growth driver, AT&T was equipped with the right data and focused prioritization to study it.

At Mobile World Congress, Markus shared the scale and metrics surrounding their AI program. Execution and adoption matter as much as the technology in question, and Markus explained that at the end of the day, with leadership’s blessing and initiation, data is what drives value, and tapping into best-in-class tools and AT&T’s robust data is what made this formula work.

“We have over 100,000 employees that have access to Ask AT&T, which is our cutting-edge generative and agentic AI platform,” he said. “It really involves taking the AI into the systems and the streams that our employees [are already] working in every day.”

Andy Markus, chief data and AI officer at AT&T, explains why they are not a pilot company.

Markus shared that AT&T now has over 3 billion production API calls in said systems and screens, with a scale of tens of billions of tokens per day on average. At this scale, it’s now about fine-tuning the language models to further increase ROI. This also fosters cost savings.

The PwC team has helped AT&T redesign the processes that AI would touch rather than bringing agentic AI into an existing legacy process with no prep—a mistake some organizations can often fall into.

Scaling at work: Agentic AI and coding

AT&T noted that they have witnessed a lot of companies across industries falling short when the risk organizations are brought in late. That’s why they work to deliver value that matters at the scale of AT&T.

“I say that vibe coding is like the death spiral of an enterprise. We’re AI-fueled coding,” Markus said. “So what you see there is a couple of things. You see a one- or two-shot approach, not an ongoing iteration. You see the output be very close to standards—so close that the human specialist doesn’t even have to make a change.

Andy Markus, chief data and AI officer at AT&T, explains how they execute on being an AI-forward company that moves fast.

“I'll give you some examples of things that we’ve done that have really changed using this approach. One of the things we do [is] create data products in a data product. It’s a cumbersome process where you bring data from source, you apply business logic, and eventually you create a data product with the right technical business metadata that you can then hydrate across or democratize across the organization. It takes about four to six weeks to build these.”

On the flip side? “Typically, using this [AI-fueled] approach, we’ve built several in 30 minutes,” Markus said. “And we’ve had architects review the code, and in some cases they’ve made zero changes.”

From one month to 30 minutes. And all the while, the potential risk of agentic AI is managed with the same rigor and standards typical of the enterprise, every step of the way. With the right guardrails in place, experimentation has room to bloom, and AT&T has room to keep scaling.

On encouraging innovation

AT&T’s business is really leaning in, and as a result, they delivered 5x in-year free cash flow impacting ROI in 2025.

Markus echoed that having the tools needed to win remains the not-so-secret sauce when it comes to encouraging innovation within an enterprise of such scale as AT&T’s. For employees to win, they need to be set up properly—with both tools and the right security.

“A great concept we have right now, we call it Permit to Learn,” Markus added. “It’s a sandbox environment where users across the company can access data that they have

access to anyway, and use really great agentic functionality and agentic tools to learn how to build an agentic workflow in an environment that’s safe and guardrailed off from production environments.”

There’s also Level Up, a program available to AT&T developers. “The Level Up program both provides education but also measurement of how we’re doing and how we can improve the process,” Markus said. “I think it really starts with the metrics that matter, and really leaning in here that [organizations] establish their own metrics that matter so that they can drive more productive adoption of AI.”

Tools and security may have received most of the credit re: successful AI adoption, but it’s clear that forward-looking leadership is the seed that every innovation grows from.

Learn more about this business success and others with PwC.