The 28th General Conference on Weights and Measures has done something quietly enormous: it killed the leap second. Effective 20 May 2027, Coordinated Universal Time will become a continuous time scale with no more one-second insertions or deletions. The maximum permissible difference between UTC and Earth-rotation time (UT1) has been widened from 0.9 seconds to 3,600 seconds — a full hour — ensuring continuity for several centuries rather than the year-to-year uncertainty operators have endured since 1972. The trigger was a looming nightmare: a negative leap second. A March 2025 workshop convened by the CCTF and IERS estimated that the probability of needing to subtract a second from UTC would reach 30% by 2035. No negative leap second has ever been applied. The resolution bluntly states it would pose "a high risk of causing anomalies and disruption to critical infrastructures that are largely unprepared," with preparation costs estimated as comparable to the Y2K millennium bug effort. The industrial case was overwhelming. IoT networks, smart grids, dispersed multi-sensor systems, GNSS-dependent transportation, energy distribution, telecoms, and space operations all depend on UTC being monotonic. Every leap second already creates a coordination headache — cloud providers like Google have historically resorted to "smearing" the extra second across hours, producing time that doesn't conform to SI units and creating inter-system inconsistency. A negative leap second, requiring every clock to skip forward, would multiply this fragility by an unknown factor in systems never tested for it. The resolution is the product of years of institutional collaboration. The CGPM's 2022 Resolution 4 set the trajectory; the ITU-R's World Radiocommunication Conference endorsed it via Resolution 655 (REV.WRC-23), which established a 2035-2040 transition window for radio dissemination systems. The 2027 date pulls implementation forward, explicitly because the negative-leap-second risk made waiting until 2035 unacceptable. Multiple international bodies — ITU-T, URSI, IEEE 1588 — submitted formal statements urging acceleration. The deeper structural concern was fragmentation. The resolution warns that delay risked user organizations abandoning UTC entirely in favor of GNSS time scales (GPS Time, Galileo System Time), which are already continuous. If enough critical systems defected to non-UTC references, the entire concept of a single international time standard would fracture. This was the real extraction threat: not a single actor siphoning value, but entropy dissolving a coordination mechanism that took decades to build. For astronomers and anyone needing precise Earth-rotation data, UT1-UTC continues to be published by IERS and disseminated via GNSS navigation messages and internet services. The information isn't lost — it's decoupled from the operational time scale. Users who need to know how fast the Earth is spinning still can; they just won't have it injected as a disruptive step function into the time standard everything else runs on. This is a rare case of global standards governance acting preemptively rather than reactively. The resolution passed before a negative leap second occurred, before systems crashed, before the fragmentation it warned about materialized. The 20 May 2027 date gives operators 11 months to update systems that still explicitly handle leap seconds — a tight but defined window. After that, UTC is continuous, and the 52-year experiment with leap seconds is over.
