Most of the semiconductor policy that moves markets arrives through the Bureau of Industry and Security's export controls — rules about what can be shipped where. But the federal government is also the largest single buyer of technology in the United States, and what it refuses to buy is a lever in its own right. A proposed rule published in the Federal Register would amend the Federal Acquisition Regulation (FAR), the rulebook that governs federal purchasing, to partially implement a statutory ban on agencies buying products and services that contain certain covered semiconductors. The compliance clock it points to is December 23, 2027.
The rulemaking is a joint effort by the bodies that own federal procurement policy. The Office of Federal Procurement Policy (OFPP), the Department of Defense (DoD), the General Services Administration (GSA), and the National Aeronautics and Space Administration (NASA) — collectively the Federal Acquisition Regulatory Council, or FAR Council — are the issuers. The statutory authority traces to a section of the James M. Inhofe National Defense Authorization Act for Fiscal Year 2023, the annual defense-policy law that has become a frequent vehicle for supply-chain security provisions.
"OFPP, DoD, GSA, and NASA (collectively referred to as the Federal Acquisition Regulatory Council, or FAR Council) are proposing to amend the Federal Acquisition Regulation (FAR) to partially implement a section of the James M. Inhofe National Defense Authorization Act for Fiscal Year 2023 which prohibits executive agencies from procuring or obtaining certain products and services that include covered semiconductor products or services effective December 23, 2027."— Federal Register, FAR proposed rule 2026-03065, source
Two words in that text deserve emphasis. The first is "partially" — the FAR Council is implementing only part of the underlying NDAA section now, which signals that the regulatory architecture is being built in stages rather than all at once. The second is the date: December 23, 2027 is the effective date written into the statute, which means the proposed rule is putting procurement language in place well ahead of the deadline rather than scrambling to meet it.
A procurement rule, not an export control
It is important to read this record for exactly what it is and not confuse it with the controls that dominate semiconductor headlines. This is a demand-side restriction on the government's own buying, distinct from the supply-side export controls administered by BIS. The mechanism is the FAR — the clause set that contractors must comply with to sell to federal agencies. Once such a clause is finalized, the prohibition does not merely cover chips bought directly; the wording reaches products and services that "include" covered semiconductor products or services. That phrasing is what gives the provision teeth and breadth: a contractor selling a finished system — a server, a network appliance, a piece of equipment — may have to attest that the semiconductors inside it are not covered, pushing diligence obligations down through the supply chain.
What the published record does not yet resolve is the operative detail every contractor will want: precisely which semiconductors are "covered." The underlying NDAA section ties the definition to specific named producers and categories of concern, and the precise scope, certification mechanics, and any exceptions are the substance that a final rule and its accompanying guidance will have to nail down. Because this is a proposed rule, it is open for public comment, and the comment period is where affected contractors, integrators, and chipmakers will argue over how far the diligence obligation reaches and how a seller can reasonably establish compliance for a complex product with thousands of components.
Why the lead time matters
The most consequential feature of this record is its timing relative to the December 23, 2027 effective date. Supply chains for electronic systems are long; the chips inside a product sold in 2028 are designed in years prior, and design wins are locked far ahead of shipment. By proposing FAR language now, the government is giving the contractor base time to qualify alternative parts, re-source covered components, and build the certification and traceability processes that a procurement prohibition demands. A rule that landed at the last minute would be unworkable; one that arrives with roughly two years of runway is a deliberate signal to the supplier base to start moving.
For the semiconductor sector, the strategic read is that procurement is becoming a parallel track of supply-chain policy alongside export controls. Where export rules reshape which markets a chipmaker can sell into, procurement rules reshape what a systems integrator can put into a product destined for federal customers. The two tracks point in the same direction — narrowing the role of certain producers in sensitive supply chains — but they operate through different agencies and different legal authorities. A company tracking only the BIS docket would miss this one entirely, even though the practical effect on a covered producer's addressable U.S. demand could be substantial.
The honest bottom line is that this is an early, partial, proposed step rather than a finished prohibition, and the parts that will determine its real-world bite — the covered-product definition, the certification standard, and any carve-outs — remain to be settled through the rulemaking process. But the direction is unambiguous and the deadline is fixed in statute. Contractors selling to the federal government, and the chipmakers whose parts ride inside those products, now have a date on the calendar and a proposed rule to read closely before the comment window closes.