Davos 2026: AI fragmentation could drive inefficiency and energy strain, warns WEF CTO

As artificial intelligence becomes deeply intertwined with global security, geopolitics, and economic competitiveness, the risk of fragmentation across countries and regions is emerging as a serious concern, according to a senior executive at the World Economic Forum.

Stephan Mergenthaler, Managing Director and Chief Technology Officer at the WEF, said that uncoordinated approaches to AI development could have far-reaching consequences beyond innovation slowdowns. Speaking to Moneycontrol on the sidelines of the World Economic Forum’s 56th annual meeting in Davos-Klosters, Mergenthaler pointed to the danger of countries building parallel and duplicative AI infrastructure.

Also read: AI-generated fake sports news poses growing risk to teams, players and fans: Study

Such fragmentation, he noted, could also intensify the energy burden associated with large-scale AI deployment, at a time when sustainability is already under strain globally. Despite these challenges, he said the larger priority for governments and businesses alike is identifying where AI can deliver the most meaningful and transformative impact.

According to Mergenthaler, companies across regions are showing strong interest in deploying AI in ways that generate tangible value, even as geopolitical and regulatory divides persist. The emphasis, he said, is shifting toward understanding sector-specific applications rather than focusing solely on the rapid evolution of AI models themselves.

He highlighted the public sector as one of the areas with the greatest untapped potential for AI adoption. Governments, he said, could significantly improve efficiency and service delivery by integrating AI into administrative and operational processes. Mergenthaler pointed to examples such as India’s investments in digital public infrastructure as early indicators of how technology-led governance can scale, expressing hope that similar efforts will emerge elsewhere.

AI has featured prominently at Davos for several years, but the nature of the conversation is changing. Instead of focusing only on technological breakthroughs, discussions are increasingly centred on implementation challenges and real-world deployment. Mergenthaler described this as a paradox: while AI’s capabilities continue to advance rapidly, many organisations struggle to unlock its full value.

Also read: OpenAI legal battle escalates as Musk demands up to $134 billion in damages

He explained that widespread adoption often requires fundamental changes to workflows and organisational design, a transition that many institutions find difficult. Identifying successful use cases and sharing lessons across industries, he said, remains a key objective for the WEF’s work on AI.

The Forum itself has begun using AI to improve how its annual meeting operates. This year, WEF introduced an AI-powered agent called Eva, developed in partnership with Salesforce, to help participants schedule meetings, discover relevant sessions, and navigate the event more efficiently. The initiative, Mergenthaler said, reflects the Forum’s broader effort to function as an “agentic enterprise” and demonstrate how AI systems can already be applied in complex, real-world environments.