The CEO Who Bet the Entire Company on an AI No One Else Wanted to Build
There's a conversation that most CEOs in telecommunications have been avoiding for years. It’s not about technology, regulation, or margins. It’s about identity: what kind of company they want to be when connectivity ceases to be a differentiator and becomes a public utility, like water or electricity.
Kaan Terzioglu, CEO of VEON Ltd., has had that conversation. The consequences of having it earlier than his peers are beginning to reflect in the numbers.
VEON operates in five economies that the world’s major tech players have long treated as second-tier markets: Pakistan, Bangladesh, Ukraine, Kazakhstan, and Uzbekistan. Over 700 million people. Young populations with limited access to formal financial services, quality education, and healthcare. In an industry term known as frontier markets, Terzioglu refers to this as his most difficult competitive advantage to replicate.
"VEON is not a traditional telco," the executive stated to Benzinga. "We are a digital services company that also provides connectivity." The statement may sound like PR until one examines the architecture behind it.
The Bet That Consultants Wouldn't Recommend
When Terzioglu talks about sovereign artificial intelligence, he's not discussing deploying language models developed in California onto local infrastructure. He’s talking about building models trained in Urdu, Bengali, Uzbek, and Kazakh, with data generated within those borders, under their governments’ regulations, and tailored to the specific needs of those economies.
This is a project that no AI lab based in San Francisco has the incentive to tackle. Global models are predominantly trained on data in English. Profitability lies in markets already paying for premium subscriptions. Languages with hundreds of millions of speakers but low per capita income are systematically underrepresented.
VEON identified this gap not as a social cause, but as a defensible strategic position. The company's digital revenues reached $759 million, a figure that indicates years of invisible work behind the transition. Such volume of digital revenue is not built on makeshift infrastructure.
What makes this move structurally different is its logic of integration. Terzioglu described AI not as a separate line of business but as a "horizontal capability" that spans fintech, education, health, entertainment, and enterprise services. This means that every product in the portfolio becomes more valuable with the same investment in the underlying model. The economics of this design resemble a tech platform more than a telecom operator.
What This Decision Reveals About the Type of Leadership That Executed It
There is a dimension to this story that financial analyses rarely touch upon, yet it is precisely the aspect I find most intriguing when diagnosing an organization’s maturity.
Positioning as a digital services company in markets where basic infrastructure is still inconsistent, where regulation changes with governments, and where macroeconomic volatility is structural, requires a risk tolerance and clarity of purpose that most boards do not sanction. The institutional pressure in any publicly traded company on NASDAQ pushes in the opposite direction: towards consolidation, predictable markets, and metrics that analysts can comfortably model.
The decision to build sovereign AI rather than simply licensing existing third-party technology is the type of choice that incurs short-term costs and yields long-term results, exactly the combination that breeds the most internal friction in an organization. It necessitates alignment among finance, technology, operations, and regulatory teams around a narrative that does not yet have a clear precedent in the industry.
This only happens when a leader has those difficult internal conversations before making public announcements. When the CEO does not postpone the debate about corporate identity to avoid discomfort in the boardroom. Organizations that achieve structural transformations do so not because they had better ideas: they do so because they had more honest conversations about their limits and their bets.
The reverse pattern, which I see more frequently in telecom companies announcing "digital transformation" for a decade without executing it, has a consistent diagnosis: the institutional ego prefers to defend the narrative of being a successful telco rather than embrace the vulnerability of building something new from within.
The Risks That the Optimism of the Announcement Does Not Cover
It would be dishonest to conclude the analysis without pointing out the real tensions this model faces.
Ukraine is simultaneously one of VEON's markets and a country in active war. Operating critical digital infrastructure in that context is not only logistically complex but also poses a continuity risk that no financial model can fully mitigate. The sovereign bet carries an implicit cost: when the sovereign is under existential pressure, so too is the company that banked on that framework.
In Pakistan and Bangladesh, regulatory stability has historically followed short political cycles. AI models trained with local data under specific regulatory frameworks can become trapped assets if those frameworks change. The digital sovereignty that is a competitive advantage today could become an operational restriction tomorrow if governments decide that locally trained data cannot exit the country under any circumstances, even for technical maintenance.
Finally, there is the risk of pure execution. Building genuine AI capability requires talent that is globally scarce. Attracting and retaining language model engineers capable of working in Urdu or Uzbek from Karachi or Tashkent, competing for salary with tech companies with access to virtually unlimited capital, presents a human resources challenge that no purpose-filled discourse can resolve on its own.
The Culture Built When the CEO Clearly Knows Who They Are
What Terzioglu is constructing at VEON has organizational implications that go beyond product strategy. When the CEO can accurately say, "we are a digital services company that also provides connectivity," that clarity is not cosmetic. It results from having resolved an internal question that many organizations of its size deliberately leave ambiguous to avoid conflict.
Companies that do not resolve this question end up investing in two contradictory directions, building teams with incompatible cultures and generating cycles of frustration that manifest as talent issues, execution problems, or speed challenges. What appears from the outside as an operational problem is almost always a symptom of a corporate identity that leadership did not have the courage to define clearly.
VEON bet on markets that no one wanted, languages that no one trained in, and a thesis that its competitors could neither validate nor desired to. If this bet scales effectively, the outcome won’t just be financial. It will demonstrate that **the culture of an organization is the natural result of pursuing an authentic purpose, or the inevitable symptom of all the difficult conversations that the leader's ego prevents them from having.
