Founded On Technology Innovation, AT&T Is Charting A Data And AI Future

Forbes

6 days ago

AT&T
How does an iconic American company that has been synonymous with technology innovation for nearly 150 years prepare to grow and thrive in an AI future? This is the question that I posed to Andy Markus, Chief Data and AI Officer at AT&T, a company that was essentially founded in 1876 when Alexander Graham Bell invented the telephone. For nearly a century and a half, AT&T has been a pioneer in technology innovation.
Markus joined AT&T in 2020, having held technology leadership and transformation positions for leading media companies including WarnerMedia, Turner Broadcasting, and Time. In his role as Chief Data and AI Officer at AT&T, Markus supports the consumer and business lines of the $122b (as of the end of 2024) company, as well as back-office functions ranging from finance to legal to HR. 'We're responsible for developing and executing the data and AI strategy and governance for the firm' notes Markus. He adds, 'The big hat we wear is execution. We work across the firm horizontally to help all parts of the business. We solve their challenges with a data and AI first mindset.' The scope of responsibility of the Chief Data and AI Office is magnified by the size and the scale of data that AT&T manages.
AT&T has a long history working with AI, dating back to pioneering work at Bells Labs, the former R&D arm, which was renowned for its groundbreaking innovations, including the invention of the transistor in 1947. Bell Labs revolutionized modern electronics and computing and played a pivotal role in the early development of AI. 'AT&T has a very rich history with AI. I like to use the line from Hamilton – 'we were in the room where it happened'. AT&T was right there when the term artificial intelligence was created' comments Markus. He adds, 'We have a rich history of technology innovation at AT&T. We recently ranked sixth in U.S. companies with AI patents and continue to turn out a considerable volume of intellectual property resulting from generative AI and agentic AI.'
As has been the case during its long history, AT&T continues to pioneer technology innovation, now using AI. 'AI is a core part of the AT&T mandate and how AT&T runs its businesses' says Markus. 'We still have the spirit of Bell Labs.' He adds, 'It's remarkable that we're one degree removed from somebody that worked with John Tukey, the legendary Bell Labs mathematician and statistician that I studied in school.' The emergence of generative AI and agentic AI in the past few years has been accelerating transformation within AT&T. Markus notes, 'We recognized that generative AI would bring AI to everyone. Instead of having AI being run exclusively by technical people, we are creating a general-purpose AI that can apply to areas where we have never used AI before.'
AT&T currently runs over 600 traditional Machine Learning and AI models in production across the firm, cutting across many lines of business. Markus explains, 'Where we were leveraging traditional or classical AI to run the business, now we're integrating every part of the firm and reimagining the things that we can do using generative AI.' Generative AI is also being employed by AT&T to help manage its data. 'Data functionality using generative AI is great for complex analytics. We are working with hard, complex, messy data sets' notes Markus. He continues, 'When we apply generative AI technology to a curated data product, the accuracy skyrockets. Generative AI technology enables us to do things that are at human level or actually exceed what can be done at a human level.' Markus adds, 'We are building on a foundation of being great with data, great with classical or traditional AI, and now great with generative AI and agentic AI. Each element builds off each other and complements each other'. He notes that AT&T has created over 2,000 generative AI use cases that have been submitted for internal reviews.
Delivering business value from AI is central to the business mandate of AT&T. 'An AI first mindset starts with understanding the business needs' notes Markus. He continues, 'Once we understand what these needs are, we work to automate processes and make the lift lighter for the development community so that they can do their work faster.' Markus adds, 'Our partners in the business are truly the experts at creating a business case. Whatever you do, you've got to integrate with existing systems. We help evaluate the cost of the solution and then we work to help understand the benefit and that's where we really work hand-in-hand our business partners.'
The AI use cases that AT&T is developing cut across the firm. Markus notes that the very first thing that AT&T did was to bring together the risk organizations of the company -- legal, compliance, privacy and security – to develop a unified approach to govern how the company should invest and execute in AI capabilities across the organization. He continues, 'We partnered with the business units to create a transformation program for this new era of AI'. AT&T has established a transformation office which reviews each use case for its business value to the organization. Markus adds, 'We prioritize our use cases and work on those that will have the most value for the company, working closely with the CFO office.'
The result of these efforts is that AI is driving business value for AT&T across business lines. In one example, AT&T has developed a complex fraud detection system. Markus explains, 'Your phone is a very expensive piece of equipment. The bad guys want to find a way to get your information that's on it.' He continues, 'To address this, we created a very complex fraud detection system with well over 30 models, both generative AI models and traditional AI models, that protect customers from fraud.' AT&T is also using AI to manage robocalls. Markus comments, 'When I started with AT&T, one of the top complaints from our customers were robocalls. At this point, by using AI we're detecting these earlier and blocking them.'
AT&T is also applying AI to deliver business value through its Ask AT&T platform. Markus explains, 'One of the areas where we've been successful in using generative AI to take human language and turn it into computer language and do complex analytics is with our Ask AT&T platform.' He elaborates, 'At the very beginning of generative AI, we saw that generative AI was going to touch all of our employees, so we created a formal AI policy.' Over 100 thousand employees and contractors now have access to the Ask AT&T platform. Markus adds, 'We are leading the pack in using AI to drive value. We do it in a very measured way across the firm.'
Another example is dispatch optimization. AT&T operates one of the largest vehicle fleets in the United States, comprising over 50,000 vehicles and over 700 million possible routes on a given day. The company has developed an AI application that optimizes the dispatch process. Markus notes, 'The benefit of the application is good for society because by being very efficient, we're saving carbon emissions. We now have over 100 million pounds of emissions saved since we started this program by reducing the miles driven. We don't always need to send technicians to homes when people call in. We've used AI to become much smarter on how we solve issues proactively, which saves a technician dispatch in many situations.'
Managing data as a business asset is core to the success of the AI transformation taking place at AT&T. 'There is an enormous amount of data that flows over the AT&T network every day – close to 900 petabytes of data that come over the network every day', says Markus. He explains, 'Our data must be safe and secure. We have this concept of a data product, which in our view is a curated set of raw data.' Markus adds, 'We need to do the right thing with how we manage our customer data to fully adhere to regulations and to drive business value for the company and our customers.'
For most organizations, and particularly century-old companies, transformation and change is seldom an easy proposition. The greatest challenges that these firms face almost always relate to the business culture and readiness of an organization to adapt. Markus notes, 'Culture is one of the driving factors for us. It is where many organizations get stuck.' He elaborates, 'We are a 150-year-old company, so inevitably there will be some parts of the business that ask whether they can really benefit from leveraging AI.'
To address some of the cultural challenges, AT&T has established AI training programs so that employees can start to understand how AI can augment their daily activities. AT&T now has employee education programs where 50,000 employees have completed AI training, and new trainings are continually being added. Markus notes that support for AI starts at the top, explaining, 'We have strong top-down support, beginning with our CEO, John Stankey. He has been a great leader, a coach and a person that evangelizes that AI is a technology we are going to embrace as a company.' While leadership from the top of the organization is essential, Markus notes that support at all levels of the company is required to ensure successful adoption of AI.
AT&T is preparing for an AI future. This entails staying abreast of the latest AI capabilities, including agentic AI. Markus explains, 'Agentic AI is connecting together a workflow that actually could be deterministic, to break down bigger problems into smaller problems that can be solved more accurately, while having the ability to take action as part of the workflow.' He adds, 'People often try to make agentic AI sound like more than it is, but it really is breaking down the big problem and smaller problems so you can solve these more accurately with the ability to take action based off of your decisions.'
Looking to the future of AI at AT&T, Markus observes, 'There's a lot of hype around whether there will be value coming from our AI investments. I think everyone's seeing that the technologies we're working with such as generative AI and the evolution of this into agentic AI are going to change things.' He comments, 'As we talk about the benefit of generative AI, for every dollar we invested, in that same year, we returned 2X ROI. This was free cash flow impacting ROI from multiple year business cases. A 2X return is going to grow to a run rate that contributes to AT&T's overall run-rate savings target of $3 billion by the end of 2027.'
Markus continues, 'What's different this time is that we don't see the wall as we have with other technology evolutions, where you had a general idea of where the where the end was going to be while you're in the middle of it. In the case of AI, I don't think we see this wall yet. The wall keeps moving, if there even is a wall. That is something that I think is super exciting to be a part of.'
He concludes, 'AT&T is an exciting place to work because the scale and complexity of our data is extremely unique. We see adoption across the board in using and reimagining how we do our work with an AI and data-driven mindset as a way to get to the next stage. We have business teams that you would never expect to be learning how to use AI that are doing so. They are chomping at the bit, knocking on our door to ask how they can be using AI to deliver better products and services to our customers. That's a great place to be!'