Post-quantum encryption standards were finalised in 2024, and major institutions have since set transition deadlines. The migration is underway and is going to take years.
The obvious question is why, given that no quantum computer capable of breaking current encryption exists or is close.
Harvest now, decrypt later
That is the answer, and it is the whole reason for urgency. Encrypted traffic can be captured today and stored indefinitely. If a capable machine is built in fifteen years, everything recorded between now and then becomes readable at that point.
For most traffic that does not matter – nobody cares about a fifteen-year-old session. For medical records, state communications, industrial designs, and identity documents, confidentiality is expected to last decades.
Data with a long confidentiality requirement needs protecting now, because the interception is happening now.
What is being replaced
Public-key cryptography specifically – the part that establishes a shared secret over an untrusted channel. Its security rests on mathematical problems a quantum computer would solve efficiently.
Symmetric encryption, which protects the data once a key is established, is much less affected. Doubling key length restores the margin.
Why migration is slow
Cryptography is embedded everywhere and mostly invisible. Protocols, certificates, hardware modules, embedded devices, code signing, and equipment installed a decade ago with no update path.
Most organisations cannot produce an inventory of where they use public-key cryptography, and you cannot migrate what you cannot find. That inventory is the actual first step and it is the step everyone underestimates.
What to do
For an individual, nothing. Browsers and messaging apps handle this and the transition is invisible.
For anyone running infrastructure: start the inventory. Not the migration – the list. Everything else depends on it, and it takes far longer than expected.