Medicare’s new payment model is built for AI, and most of the tech world has no idea

There is no governmental mechanism to pay for an AI agent that monitors a patient between visits, calls to check in, coordinates a housing referral, or makes sure someone picks up their medication. ACCESS creates that mechanism for the first time.

What happened

There is no governmental mechanism to pay for an AI agent that monitors a patient between visits, calls to check in, coordinates a housing referral, or makes sure someone picks up their medication. ACCESS creates that mechanism for the first time. The development sits squarely in the tech and AI cycle that our editors have been tracking this week, touching on agent, ai, first, model.

Why this matters

AI infrastructure spending is the defining capex story of 2026, with hyperscalers and frontier labs deploying tens of billions into chip supply, model training, and enterprise deployment. Each milestone like this feeds directly into the compute-supply race that's increasingly setting the tone for enterprise software pricing, equity-market leadership, and the global semiconductor trade.

The bigger picture

Throughout 2026, frontier AI labs have been on an unprecedented spending arc — Nvidia, Anthropic, OpenAI, and the major hyperscalers committing tens of billions to compute, model training, and enterprise deployment partnerships. The pace shows few signs of slowing: each new earnings cycle has revealed larger commitments than the last, and the gap between the labs that can afford frontier compute and everyone else continues to widen.

What to watch

The next data points to watch: Q2 earnings from the largest labs and hyperscalers, AI-chip ASP trends, and whether enterprise budgets for AI deployment are accelerating or hitting a digestion phase. Each set of guidance from the major players will telegraph the direction for the rest of the cycle.

Originally reported by TechCrunch. Read the original report for full context.