The gap between AI’s ubiquity and its measurable commercial contribution is widening. A global survey of executives by Harvard Business Review Analytic Services found that while 59 per cent of organisations have artificial intelligence in production, fewer than 30 per cent see any impact on new revenue streams, suggesting deployments deliver efficiency rather than transformation. Uber’s chief operating officer this week acknowledged that coding assistants boosted internal productivity by 25 per cent but admitted “no direct link” to better consumer features. In London, where services dominate, the disconnect between hype and tangible return is reshaping livelihoods.

For many white-collar workers, disruption is acute but not yet apocalyptic. A British translator recounted training her own AI replacement, yet OpenAI’s Sam Altman told a Sydney conference he had been “delighted to be wrong” about the scale of early job destruction. New research suggests remote work, not generative AI, may be the larger factor behind the sharp decline in entry-level hiring across four Anglophone economies; when controlling for both, the AI effect often lost statistical significance. Meanwhile, algorithmic bias is climbing the corporate agenda in Brazil and beyond, as Stanford’s AI Index recorded 362 documented incidents in 2025, up from 233 the year before, intensifying demands for explainability in credit, hiring and customer service systems.

The advertising world is bracing for structural upheaval. Google’s new agentic AI can plan, book and purchase on a user’s behalf, potentially bypassing brand websites entirely, a shift one ad executive called “a structural change.” Google’s chief executive, Sundar Pichai, admitted that an AI-generated search result was “more opinionated than it should be,” underlining the tension between conversational ease and factual rigour. Meanwhile, OpenAI is building an ad engine around conversational intent rather than keywords, which early data from Similarweb suggests could challenge Google’s search empire. “The user’s intent evolves during the conversation itself,” noted one analyst, describing a fundamental rewrite of the digital advertising model.

The financial and cultural foundations of the AI boom are increasingly fragile. Executives in Mexico warn that generative AI is “beginning to fracture” under unsustainable processing costs and a lack of large-scale solutions, while youth scepticism grows. Calls from Washington to tax AI are gaining traction, with advocates arguing that without policy intervention the technology will create a permanent underclass. In Sweden, 84 per cent of financial firms report employees using generative AI, but only 22 per cent have systems in production, illustrating the chasm between experimentation and operational transformation. The year ahead will test whether AI can evolve from a productivity booster into a true value generator, and whether societies can negotiate its rewards more fairly.