With over 143 million subscribers in the U.S. and a global footprint spanning over 150 countries, Verizon stands as one of the world’s largest telecommunications and technology companies. As AI reshapes industries, Verizon is reimagining itself once again, embedding AI across every layer of its operations while upholding trust, transparency, and accountability.

The first part of the three-part series explored how Verizon’s data and AI leadership is evolving, detailing the architecture of its AI strategy and the ways applied AI is enhancing customer and employee experiences. Part two covered how the company aligns AI initiatives with C-suite priorities, and how it builds a strong data foundation while scaling adoption through partnerships and workforce empowerment.

In this final part of the series, Mano Mannoochahr, Chief Data Analytics and AI Officer at Verizon, sits down with Jeff Burlin, Senior Director of Executive Engagement Programs at Twilio, to discuss how Verizon is managing risk, scaling enablement, and exploring the frontiers of agentic AI, all while preparing its people, platforms, and principles for the next era of transformation.

Training for responsible AI

Rolling out AI tools across a massive workforce requires both empowerment and vigilance. Mannoochahr emphasizes a dual approach: “We come at it both from a defense as well as a productivity perspective. You have to make sure that people know AI’s power and its limitations.”

Every Verizon employee using AI tools undergoes mandatory training before access is granted. “I had to take the training when I joined because I’m new at the company,” he notes, underscoring that accountability starts at the top. These sessions cover not only tool usage but also key concepts such as bias, hallucination, and responsible decision-making. “AI may be giving you answers, but it doesn’t mean that we don’t apply expert judgment to that.”

The responsible AI framework: Governance at scale

To ensure ethical and transparent AI adoption, Verizon has built a Responsible AI Framework, which Mannoochahr says is playing a critical role.” The framework governs everything from data privacy and transparency to bias detection and technical robustness,” he says. The framework also integrates peer reviews, legal evaluations, and ethical oversight before any model goes into production.

These values are encapsulated in Verizon’s VZI Principles, a set of foundational commitments that blend business ethics with social responsibility. “Since we’re a consumer-oriented company, we’ve got to have respect for human rights and responsibility to our connections in the communities we operate in,” Mannoochahr adds.

Agentic AI in action

Verizon’s innovation momentum is already moving toward agentic AI. “This is the recent leg in the journey to the maturity of AI that all of us are seeing. In fact, one of my team members told me, we’ve got hundreds of agents being developed already,” Mannoochahr reveals.

These agents are not abstract prototypes; they are already in production, he adds. “We’ve got customers who may be shopping for certain offerings, and the agents in the back are assembling the right products and the right solution that they may need,” he says.

Still, Mannoochahr emphasizes a human-in-the-loop approach to balance innovation with accountability. “Now AI is telling us how to do something; it offers to do the work. We’ve got to be able to apply that responsibly and make sure we’ve got a human in the loop on critical things.”

Guardrails are essential, he stresses: “If you leave too much ambiguity, they could go awry, and the direction they may take may not be congruent with what we want them to be doing.”

The role of data and context

As agentic AI advances, the importance of institutional knowledge becomes more apparent. Mannoochahr points out that much of this knowledge lives in employees’ minds rather than structured systems. “How do we ensure that these knowledge assets are now finding their way to the same set of standards and infrastructure that we use to manage our structured data?” he asks.

Verizon’s evolving data infrastructure aims to integrate procedural, unstructured, and contextual knowledge, giving AI agents the depth they need to perform effectively while adhering to organizational rigor and compliance.

The humor and humanity behind AI governance

Even as AI grows more autonomous, Mannoochahr jokingly says, “I can’t wait to send these AI agents to compliance training because today, as humans, we have to put up with it. These agents better learn our code of conduct as well.”

Behind the humor lies a serious truth: that ethical alignment will be just as important for AI systems as it is for people. “They’re going to be doing work, and they have to know how we operate as a company,” he says.

When asked about the future, Mannoochahr shares that AI will redefine industries, create trillion-dollar enterprises, and fuel global economic growth. “This is our opportunity to do a better job than our competition with this new technology and improve our customers’ lives as well as our business,” he concludes.

CDO Magazine appreciates Mano Mannoochahr for sharing his insights with our global community.