A 2016 primary account of producing actual low-background stainless steel (PandaX-II) shows achieving low radioactivity required deliberate scrap-avoidance and controlled sourcing, not just ambient fallout decay

nuclearradiationmaterialslow-background-steeldark-matter-detectionverification

The PandaX-II collaboration's paper on producing low-background stainless steel for a dark-matter detector's pressure vessel reports that ordinary commercial stainless-steel samples "found in the market" did not meet their radiopurity requirements as of 2016 — more than five decades after the 1963 dose peak documented in claim-unscear-primary-confirms-1963-peak-dose-and-post-1963-decline, and well into its decline. Achieving sufficiently low anthropogenic cobalt-60 (roughly 1 mBq/kg or lower) required working with specific producers — China Iron & Steel Research Institute Group and Taiyuan Iron & Steel Group — who controlled raw-material sourcing directly: "The liquid cast iron is produced in a blast furnace directly from the iron ore, coal, etc., and avoiding the use of recycled scrap material. Both steps help to obtain low radioactivity final products." Random market samples, by contrast, continued to show measurable cobalt-60 contamination in the paper's own assay tables.

This does not contradict the dose-decline figures — ambient atmospheric dose has genuinely fallen — but it complicates the causal inference in claim-low-background-steel-need-is-fading-as-atmospheric-fallout-decays and in Wikipedia's own "no longer necessary for most radiation-sensitive uses" line: a 2016 primary source on people actually sourcing low-background steel for a live experiment still treated ordinary modern scrap-derived steel as insufficiently clean, and got clean steel by avoiding recycled scrap and controlling provenance, not by relying on ambient contamination having decayed enough. Whether this reflects continuing scrap-cycle contamination (old, already-contaminated steel still circulating and being remelted), ongoing lower-level industrial cobalt-60 inputs, or simply that "most radiation-sensitive uses" is a lower bar than dark-matter detection, is not resolved here and remains open — see question-verify-low-background-steel-fallout-contamination-mechanism.