The End of Leap Seconds - The 2035 Decision and Its Technical Impact

What Leap Seconds Are and Why They Were Inserted

Leap seconds are adjustments inserted into UTC to keep it within 0.9 seconds of UT1, the time scale based on Earth's actual rotation. Since UTC was introduced in 1972, 27 leap seconds have been added through the end of 2016. They are inserted as 23:59:60 immediately following 23:59:59 on June 30 or December 31, an unusual second value that does not appear in normal timekeeping.

The need for leap seconds arises because Earth's rotation is not constant. Tidal friction is gradually slowing Earth's spin, lengthening the day by about 2.3 milliseconds per century in the long run. Short-term variations from tectonic motion and atmospheric circulation can speed it up, so the interval between leap seconds is not fixed. From 1972 to 1979 they came annually; since 2017 there has been none, reflecting a temporary acceleration of Earth's rotation.

The 2022 CGPM Decision - International Agreement to Abolish

In November 2022, the 27th General Conference on Weights and Measures (CGPM) adopted a resolution to halt the insertion of leap seconds by 2035. The decision reflects a judgment that the operational risks leap seconds pose to communications, finance, and navigation now outweigh the value of keeping civil time aligned with Earth's rotation.

The resolution allows the gap between UTC and UT1 to grow beyond one second and leaves open the possibility of a future, larger correction (for example, a leap minute) in decades or centuries when the gap becomes substantial. The International Telecommunication Union (ITU) is developing the implementation details by 2035, including how the existing leap-second framework will wind down operationally.

The IT Headache - Why 23:59:60 Breaks Things

Leap seconds are problematic for software because the value 23:59:60 is not legal in standard time representations. Most software assumes 60 seconds per minute and cannot represent or process the 61st second. The 2012 leap second exposed a Linux kernel bug that crashed Reddit, Mozilla, Foursquare, and other major services in a coordinated incident.

Air traffic control must carefully manage the difference between GPS time (which does not include leap seconds) and UTC across leap-second boundaries. Financial markets timestamp trades at high precision, and a one-second discontinuity can scramble trade ordering for the duration of the leap second. These risks are why most large operators have lobbied for abolition for years.

Leap Smearing - The Workaround in Use Today

The dominant workaround for leap seconds is "leap smearing," introduced by Google in 2008. Instead of inserting a single 23:59:60, leap smearing stretches each second slightly during a 24-hour window centered on the would-be leap. Each second becomes 1/86,400 of a second longer (about 11.6 microseconds), spreading the one-second adjustment over a full day.

Amazon (AWS) and Microsoft (Azure) use similar smearing techniques, though the exact start time and duration differ between providers. Comparing timestamps across cloud providers near a leap second can produce small but real discrepancies. After 2035 abolition, leap smearing itself will become unnecessary, simplifying interop between cloud platforms.

Life After 2035 - What Will Change

Once leap seconds stop, UTC will effectively run at a fixed offset from TAI (currently 37 seconds). Variations in Earth's rotation will be reflected only in UT1, leaving UTC unaffected. Astronomers and certain navigation systems will continue to use UT1, but everyday IT systems can rely solely on UTC without further adjustment.

Over the long run, the gap between UTC and solar time (UT1) will accumulate. At current rates of Earth's deceleration, the gap will reach about one minute in 100 years and roughly one hour in 1,000 years. These shifts unfold over generations, far slower than any human can perceive. The CGPM decision essentially passes the problem of long-term astronomical alignment to future generations to handle, while removing the operational hazard for our era.