FCC kills the gigabit goal, and 100/20 becomes the planning floor

The FCC finalized its latest broadband deployment report on August 14 and abolished "without replacement" the long-term goal of 1,000Mbps down and 500Mbps up that the agency adopted in 2024 ^[1][2]. That leaves 100Mbps down and 20Mbps up as the only operative federal speed definition, and it is the benchmark by which the agency judges whether deployment is reasonable and timely ^[3].

The 2024 report did two separate things. It raised the benchmark from 25/3 to 100/20 ^[4], and it set 1Gbps/500Mbps as a "guidepost for evaluating our efforts to encourage deployment" ^[3]. Chairman Brendan Carr proposed abolishing that guidepost last year ^[6]. The order's stated reasoning is that a long-term goal "is not mentioned in Section 706 and could appear to violate our obligation to conduct our analysis in a technologically neutral manner," and that "it is impossible to predict long-term technological developments and the evolution of consumer preferences" ^[2]. The reports themselves are mandated by Section 706 of the Telecommunications Act ^[7].

For anyone sizing a product against the American median, the number that matters is the 20 in 100/20. The retired goal was 10 times the download benchmark and 25 times the upload benchmark ^[10]. The surviving floor is 5:1 asymmetric where the goal was 2:1 ^[11]. At 20Mbps up, a single 1GB upload takes roughly 6.7 minutes ^[13]. Anything you ship that assumes headroom on the return path, including video capture, continuous backup, remote development, screen sharing at high frame rates, and camera-side inference offload, is now designed against a federal definition that no longer contains an upstream ambition. Fiber networks have routinely offered gigabit speeds for years ^[8], so the capability is real in the market; what changed is the target the regulator uses to describe adequacy.

The same report declares current deployment levels acceptable ^[1] and states that speeds are up, competition is up, and prices are down ^[5]. Ars Technica notes that some of those findings draw on industry lobby group material and that no adverse research from advocacy groups was used ^[5]. The FCC also said comments from the cable group NCTA and the wireless group CTIA "support our reasoning for abolishing the long-term goal" ^[12]. Last year the Carr FCC said the gigabit goal "may be unreasonably prejudicial to technologies such as satellite and fixed wireless that presently do not support such speeds," though the new report says that is not why it is eliminating the goal ^[9].

That matters for subsidy-funded buildouts, because a lower stated ambition makes slower technologies look sufficient. Techdirt reports that Republicans have redirected large portions of the $42.5 billion set aside for broadband access toward Elon Musk's Starlink and away from faster fiber networks and cooperatives ^[14]. Techdirt also notes it took more than a decade of pressure against ISP lobbying before the FCC adopted 100/20 at all ^[15].

Three things to watch. Whether the next Section 706 report leaves 100/20 untouched, since with no long-term goal above it, the benchmark is the entire federal position. Whether state broadband offices continue to treat 100/20 as the served threshold when awarding money, which determines how many addresses get a fiber build versus a satellite terminal. And whether consumer upstream tiers move at all over the next two years now that no federal guidepost points past 20Mbps, because that is the variable that decides which product behaviors are viable by default and which need a local buffer and a retry queue.