By: Guy Lupo

Not long ago, telecom operators were asking whether they should adopt artificial intelligence (AI). Today, the question is very different.

AI is already here. It is embedded in network operations, customer service, security functions, software development and business processes. New AI agents, models and tools appear seemingly every week, while employees and suppliers continue to experiment with new capabilities. The challenge facing operators is no longer adoption. It is visibility.

A useful analogy comes from an unlikely source: Pokémon.

The current AI landscape often feels like a world filled with Pokémon. They appear everywhere. New ones emerge constantly. Different teams are using them for different purposes. Some are approved and governed. Others may not be.  Leaders are left with the hard part: finding them all, understanding what each one can do, and building a complete, living inventory.

For telecom operators, that question has significant consequences. Every AI agent, model, or application potentially introduces new decisions, new data dependencies, and new operational risks. Before organizations can truly benefit from AI at scale, they need to know what AI exists within their business, how it behaves, and whether it can be trusted.

This is becoming one of the defining challenges of the industry's journey toward autonomous operations.

The evolution of AI leadership

The rise of AI is also reshaping leadership within telecom organizations. A few years ago, Chief Data Officers often found themselves responsible for managing data platforms while struggling to demonstrate business value. AI initiatives, meanwhile, frequently sat within isolated AI programs, IT teams, or specific business functions. Those boundaries are rapidly disappearing.

Today's AI systems depend upon access to data, applications, network capabilities, and security controls. An AI agent cannot deliver meaningful outcomes if it operates as a disconnected tool sitting outside the operational environment. It needs to interact with enterprise systems, understand data, and execute actions across the organization.

This convergence is creating a new generation of leadership roles. Chief AI and Data Officers are increasingly being positioned closer to executive decision-making and are assuming broader responsibilities than technology enablement alone. Their remit is expanding to include governance, risk management, security, sovereignty, and operational accountability.

In some organizations, this evolution is already producing a new title: Chief AI and Trust Officer. As AI becomes more influential in operational decision-making, leaders will inevitably be asked the same question by boards, regulators and executives: Can we trust it to run our business? Answering that question requires more than technical expertise. It requires evidence, governance and accountability.

Automation vs. autonomy

The telecom industry has spent decades pursuing automation. Automated workflows, policy-based controls, and intelligent operations have become familiar parts of modern network management. Autonomy represents something different. Automation focuses on executing predefined actions. Autonomy requires systems to make decisions within established boundaries and adapt to changing circumstances while remaining aligned with organizational objectives.

Many organizations describe advanced AI as automation on a larger scale. In practice, true autonomy requires a fundamental shift in how decisions are delegated to machines. The industry is moving from models that rely on humans reviewing every action – “human in the loop” – toward environments in which humans establish objectives, policies, and controls while autonomous systems determine how those outcomes are achieved – “human in charge of the loop.” That transition cannot occur without trust.

Organizations must trust the data used to train AI systems, the models generating recommendations, and the behavior of agents interacting with critical infrastructure. They must also trust that they can intervene, audit, and explain decisions when required. Without that trust, AI remains confined to pilots and demonstrations. With trust, it becomes possible to scale autonomy across complex operational environments.