Zoom out from the quarter and the most consequential telecom-subsidy document of the year is not a rate ruling but a reframing. On June 5, 2026, the FCC adopted a Notice of Proposed Rulemaking, FCC 26-35, under WC Docket Nos. 26-96 and 10-90, to overhaul the High-Cost Program — the part of the Universal Service Fund that pays carriers to build and run networks where the economics otherwise do not work. What makes it a long-bet story rather than a housekeeping item is the rationale the Commission attached: it is positioning rural broadband subsidies as the infrastructure layer beneath the AI buildout. Comments are due August 4, 2026, with reply comments due September 3.
For a capex analyst, the interesting word in the NPRM is "predictable." Carriers do not deploy fiber or fixed wireless into high-cost areas on the strength of a single year's support; they deploy against a multi-year recovery curve, and the discount rate they apply to that curve is a direct function of how predictable the subsidy stream looks. An NPRM that promises a more efficient and effective High-Cost Program for years to come is, in capital-allocation terms, an attempt to lower the perceived risk premium on rural network investment. Whether it succeeds depends on the rules that follow, but the intent is a capex-curve intervention.
"Ensuring a predictable High-Cost Program for years to come--call it High-Cost Modernization--will provide continuing support for our Build America Agenda, supercharge American leadership in Artificial Intelligence (AI) by efficiently supporting the broadband-capable networks upon which AI-enhanced applications and services will be delivered and accessed, and will help accelerate the transition to Internet Protocol (IP) networks."— Federal Register, FCC Proposed Rule, June 5, 2026, source
The AI framing is a budget argument
Read literally, the line about supercharging American leadership in AI by supporting broadband-capable networks is a policy aspiration. Read as a budget argument, it is something sharper. The Universal Service Fund has spent years under pressure over its contribution mechanism and its long-run sustainability. Binding the High-Cost Program to the AI buildout reframes it from a legacy rural-telephone subsidy into critical national infrastructure for the technology priority of the decade. That reframing matters for one practical reason: programs framed as AI infrastructure are easier to defend, and harder to cut, than programs framed as rural-phone relief. For carriers that draw on the fund, the durability of the subsidy is the single biggest input to whether the buildout pencils out.
All-IP is the cost-side story
The other half of the NPRM is the all-IP framing, and that is a cost-side story for the carriers. The transition to Internet Protocol networks lets operators retire parallel legacy plant, consolidate operations onto a single packet-based core, and stop spending to maintain TDM equipment that the subsidy historically helped support. A High-Cost Program redesigned around IP networks is implicitly a program that stops paying to keep old technology alive and starts paying for capacity that AI-era traffic will actually use. For a long-horizon investor, the signal is that the recurring opex base of a subsidized rural network should fall as the IP transition completes, even as the capex needed for capacity rises. Net of those two movements is the number that determines the program's true cost trajectory, and the NPRM is the first place to look for how the Commission proposes to handle it.
There is also a competitive-structure consequence to the all-IP framing that is easy to miss. A subsidy program organized around IP-capable networks implicitly favors operators who have already made, or can quickly make, the transition, and disadvantages those still running legacy plant they cannot economically retire. Over a multi-year horizon, that tilts the field toward better-capitalized carriers and toward fixed-wireless and fiber builders whose architectures are natively IP, while pressuring small rural incumbents whose recovery curves were built on the old support model. The modernization framing is therefore not neutral as to who builds rural networks; it is a thumb on the scale toward the operators best positioned for an AI-era traffic profile. That redistribution is the kind of slow, structural shift a capex analyst should be tracking across the life of the docket rather than reacting to in any single quarter.
What to model, and what to discount
The discipline here is to separate the framing from the mechanism. The framing — Build America Agenda, AI leadership, the IP future — is the rhetoric that justifies the program. The mechanism is whatever the eventual rules do to support amounts, eligibility, and the formulas that convert deployment into dollars. An analyst should treat the framing as a signal of durability and direction, and reserve judgment on the actual capex impact until the support mechanics are specified. The risk is the familiar one with modernization NPRMs: efficiency language can mean a larger, more predictable program, or it can mean a smaller, better-targeted one. Both are "efficient." Only one expands the addressable buildout.
The honest framing for now is that this is an option, priced in policy rather than cash. The Commission has signaled that it wants the High-Cost Program to outlive the political cycle and to serve as the broadband substrate for AI-enhanced services. That lowers the perceived risk premium on rural network investment at the margin, which is real value to carriers even before a single rule changes. But it is the comment cycle closing August 4 and the rules that follow that will turn the framing into a capex number. Until then, treat the NPRM as what it is: a long-dated signal that the subsidy is being recast as infrastructure, and a reason to read the next round of this docket with the capex model open. The patient read is to watch for two things in the follow-on rules — whether the support formulas reward sustained capacity rather than one-time deployment, and whether the program's funding base is given the durability the rhetoric promises. Those two answers, not the AI framing itself, will decide whether the modernization expands the buildable footprint or simply reshapes how a fixed pool of support is divided.