The world’s breakneck embrace of artificial intelligence has pushed global IT spending to a projected $6.31 trillion, yet the current boom increasingly resembles a tower built on uneven foundations. Viewed from the trading floors of Wall Street and the boardrooms of Silicon Valley, the most vivid symptom of this new arms race is ‘tokenmaxxing’ — a corporate flex in which firms boast about the number of AI tokens they consume as a proxy for technical sophistication. Visa doubled its token usage from one trillion to two trillion in a single month, while JPMorgan and Disney have reportedly installed internal dashboards to track individual employees’ AI usage. The metric has become a currency of ambition, even if its link to genuine productivity remains unproven.

Across the Atlantic, European and North American technology buyers are being forced to rethink their sourcing strategies. A Bain & Company survey of 280 executives reveals that three in four expect AI to reshape their industries, and spending on technology services is set to rise modestly even as overall IT budgets stay flat. Customers are clamouring for integrated security, cloud and data solutions that embed AI, compelling services providers to bet big on automated delivery and platform-based models — or risk irrelevance.

In Mumbai and Bengaluru, the message lands with an edge. India’s IT services titans, the back office to the world, are trapped in a margin squeeze even as overall services growth ticks along at a modest three to four percent. Gartner analysts note that hyperscaler spending on networking and large language model infrastructure is inflating headline figures globally, but routine commoditised work is being pulled in-house by chief information officers who are redirecting internal teams toward strategic AI. The salvation for Indian firms, Gartner suggests, lies in managed services: filling the labour gap left by this shift and embedding agentic automation into their own offerings.

Yet from Accra to Abidjan, a far more fundamental friction is emerging between adoption and mastery. Africa is entering the AI revolution with palpable ambition — the African Union adopted its first continental AI strategy in July 2024, and UNESCO estimates the technology could add $1.5 trillion to the continent’s GDP by 2030. But the continent is using AI faster than it is building the data centres, sovereign models and computing capacity that make adoption meaningful. Without local infrastructure, much of the value risks being captured elsewhere.

The same asymmetry appears inside specific industries. A SITA survey of 70 airlines and 370 airports worldwide highlights how Middle Eastern and African carriers are aggressively increasing IT spending to boost resilience against high fuel costs and supply chain snarls. In the region, every single airline surveyed raised its IT budget. Airports, however, are lagging badly, creating a disconnect that prevents carriers from realising the full returns of their digital transformation investments. It is a microcosm of the global tension: pouring money into AI and digital tools is not the same as constructing the layered, integrated environment in which those tools can deliver.

Analysts in London caution that the next phase of the AI cycle will be unforgiving. As token-based vanity metrics give way to demands for measurable returns, the winners will be those who close the gaps — between hardware and mastery, between boardroom hype and operational reality, and between the front-end glamour of AI adoption and the unglamorous work of rewiring organisations. The race is not simply to spend, but to integrate, and that race remains wide open.