AI Data Centers Demand a New Reciprocal Compact with the Grid

The top five US hyperscalers are projecting a combined capital expenditure of $660 billion to $725 billion for 2026, nearly double their 2025 outlays, as the AI build-out shifts from software to physical infrastructure. Microsoft is guiding for roughly $175 billion in adjusted capital expenditure for both FY2026 and FY2027, with two-thirds of quarterly spend going to short-lived assets like CPUs and GPUs and the rest to long-lived data center infrastructure, and it added 1 gigawatt of capacity in Q3 FY2026, doubling its global footprint in two years. Amazon AWS has raised its 2026 capex guidance to approximately $220 billion, with CEO Andy Jassy saying AI capacity is expected to remain constrained through 2027 and contracted demand extending into 2028. Meta saw profit drop 14% in Q2 2026 despite a 28% revenue increase as its build-out, including a 1 gigawatt data center in Ohio and a Louisiana facility that could scale to 5 gigawatts, compressed margins, while Alphabet raised its 2026 capex guidance to as much as $205 billion and its Google Cloud backlog more than doubled year-over-year to $240 billion. The Stargate joint venture involving Oracle, OpenAI and others targets up to $500 billion in infrastructure investment by 2029, and Oracle's FY2026 capex reached $55.7 billion, more than doubling from the previous year.

PG&E announced a new phase of its community microgrid program backed by US$73 million in funding commitments, disclosing fresh grant agreements and a second round of awards for local clean energy resilience projects. The new funding extends the utility's community microgrid efforts into additional regions that have faced recurring outage and reliability concerns. PG&E operates a large regulated utility through Pacific Gas and Electric Company, supplying electricity and natural gas across northern and central California, and expanding community microgrids ties directly into how this US$29.5b operator manages reliability for its service territory. The clearest early signal to track is how many of the funded microgrid projects actually reach construction and successful operation on schedule, with PG&E securing timely cost recovery approvals from regulators for the roughly US$73 million already authorized and future tranches that could follow this template.

Caterpillar announced on September 15 that its self-driving haul trucks at a Luck Stone quarry are moving large tonnage, with expansion to two more Virginia quarries, Boscobel and Bealeton, including the first autonomous deployment of Cat 775 trucks. The company disclosed no order value, truck count, or profit contribution from the autonomy milestone, and autonomy was named on the earnings call only as a source of higher SG&A and R&D expenses inside Resource Industries. In the second quarter of 2026, revenue reached $20.5 billion, up 24% year over year, with adjusted EPS of $8.17, up 73%, while backlog climbed to $72 billion, up roughly $35 billion versus a year earlier. Power & Energy is the engine, with segment sales up 17% to $8.2 billion and power generation sales up 72% on large gensets and turbines for data centers, and gas prime orders extend toward the back half of 2028 and into 2029. Caterpillar trades at $798.51 against a consensus analyst target of $975.61, a trailing PE of 34x, and full-year tariff costs of around $2.2 billion for 2026, with expected IEEPA tariff recoveries of approximately $400 million.