Summary
- Hong Kong’s Communications Authority granted the prior consent required for SmarTone to cease 3G service on 9 October 2026.
- The regulator says a very low proportion of SmarTone customers will be affected.
- SmarTone must maintain satisfactory 3G service until the cessation date.
- More than 99% of mobile users across Hong Kong already use 4G or 5G, a territory-wide statistic rather than SmarTone’s own customer ratio.
- Older devices, voice fallbacks and specialised equipment make the remaining migration work more consequential than its small denominator suggests.
Who still bears the migration cost before 9 October? Regulatory consent gives SmarTone a fixed end date, but it does not make the remaining users or devices disappear. Until cessation, the operator still has to deliver satisfactory 3G service and help the affected edge move.
The Communications Authority granted the prior consent required under Special Condition 10.4 of SmarTone’s licence. The company may cease 3G on 9 October 2026. “May cease” identifies a future authorization; it is not evidence that the network has stopped.
The regulator describes the affected proportion of SmarTone customers as very low. It separately says more than 99% of Hong Kong mobile users already use 4G or 5G. Those are different denominators. The second figure covers the territory and cannot be substituted for a SmarTone-specific base.
Small in number does not mean simple to migrate
A subscriber can own a 4G-capable handset and still depend on 3G for some functions if the device, software version, SIM configuration or service plan does not support the operator’s newer voice arrangements. Other affected connections may sit inside alarms, payment terminals, telemetry devices or backup communications that receive less attention than a consumer smartphone.
The regulator’s announcement does not publish a device inventory. It therefore supports neither a claim that no customers remain nor a precise estimate of replacement cost.
Migration is more than selling a new handset. The user may need compatibility confirmation, a software update, a replacement SIM, plan changes or testing of voice and emergency functions. A business device may require a technician and operational downtime.
That creates an allocation question. Customers may pay for newer equipment, operators may provide targeted offers, and equipment owners may absorb installation work. The consent does not, by itself, state that every migration is free.
Service quality remains an obligation until the date
A network retirement can create a perverse final-year incentive: investment naturally shifts to newer technology even though legacy users still rely on the old layer. The authority’s requirement for satisfactory 3G service through cessation constrains that decline.
SmarTone cannot treat the authorization as permission for an unmanaged fade-out. Coverage, fault handling and ordinary service need to remain adequate until 9 October. Clear notices also matter because a fixed date has little protective value if an affected user cannot identify that a device depends on 3G.
The transition clock gives customers time, but it also gives the operator a measurable responsibility. Evidence should include direct notification, compatibility checks, migration offers, complaint handling and the resolution of persistent legacy cases.
The territory-wide adoption figure shows why retirement is rational: keeping a complete radio layer for a shrinking group consumes spectrum, energy, equipment support and engineering attention. It does not decide how the final group should be treated.
The spectrum benefit follows a clean migration
After cessation, spectrum and operational resources can be redirected to newer services. The announcement confirms the retirement permission but does not quantify the spectrum that will be reused, its future band plan or the capacity improvement users will receive.
Those benefits should be measured after the shutdown rather than presumed from consent. Useful evidence will include the actual cessation, the handling of exceptions, changes in newer-network coverage or capacity, and whether customers experience gaps in voice or device operation.
Hong Kong’s move is a lifecycle decision, not an instantaneous technical event. More than 99% territory-wide adoption makes the old network economically marginal; the remaining very low SmarTone share makes the support task bounded. Neither makes it optional. The success test is whether the operator reaches 9 October with customers migrated and satisfactory service intact, then reuses the retired layer without creating a hidden connectivity loss.