Post-Quantum Crypto Is Coming: Who Pays for the Biggest Key Migration in Decades

Archi4 min read

cryptographybudgetsecurity

Contents
  1. The Invoice Nobody Has Priced Yet
  2. The Double-Key Years
  3. Certificates Are the Long Pole
  4. Where the Hardware Bill Lands
  5. The Inventory Problem Costs More Than It Looks
  6. Code Signing and the Trust Chain
  7. What to Watch in the Next 18 Months

The Invoice Nobody Has Priced Yet#

Post-quantum cryptography is usually framed as a technical problem: swap RSA and elliptic curve keys for lattice-based ones, update the libraries, ship the firmware. The money question is harder. Cryptographic migrations touch every certificate, every hardware security module, every signing pipeline, and every device with a shelf life longer than the new algorithm's rollout. Someone has to pay for the overlap, and the overlap is where the cost lives.

The forecast here is less about which algorithm wins. NIST's picks (ML-KEM for key exchange, ML-DSA and SLH-DSA for signatures) are settled enough to plan around. The open question is where the dollars surface in budgets over the next three to five years. Expect them in three places: certificate infrastructure, embedded hardware, and the labor of inventory itself.

The Double-Key Years#

The most predictable cost is the transition period, when systems must run classical and post-quantum keys side by side. Hybrid key exchange, where a connection negotiates both, is the default recommendation for a reason: it protects against a future quantum break without betting everything on unproven code. But hybrid means bigger handshakes, larger certificates, and more compute per connection.

Those costs are small per transaction and enormous at scale. A large certificate authority or cloud load balancer processing billions of TLS handshakes a day will feel larger keys as CPU and bandwidth. Operators I have spoken with describe the number in the low single digits of percentage points of infrastructure spend, but that is a guess made before any of them have run production traffic. The first credible public figures will come from the big CDNs and cloud providers, likely as a footnote in an earnings call rather than a press release.

Certificates Are the Long Pole#

Certificate lifetimes have been shrinking for years, and that helps. Shorter lifetimes mean a faster path to retiring classical keys once post-quantum ones are trusted. The cost is not the certificate itself but the automation that renews it. Any organization still rotating certificates by hand will discover the real bill: the tooling project they deferred, now mandatory.

Where the Hardware Bill Lands#

Embedded devices are the expensive problem. A smart meter, an industrial sensor, or a payment terminal designed in 2024 may still be in service in 2044. If it cannot run post-quantum algorithms in firmware, it becomes a liability with a decade of depreciation left on the books.

This is where budgets get real. Replacing field hardware is a capital expense with a schedule attached, and nobody wants to replace working equipment early. The forecast: expect a wave of firmware updates for devices with enough flash and RAM, a longer tail of cryptographic gateway workarounds for those without, and a smaller set of forced replacements concentrated in high-value systems (payment infrastructure, identity tokens, code signing hardware).

The money follows the risk model. Systems that must stay secure for twenty years get fixed first. Systems that rotate every three years get fixed by attrition. The awkward middle, devices with a ten-year life and no upgrade path, is where the largest unplanned spend will appear.

The Inventory Problem Costs More Than It Looks#

Before any algorithm ships, an organization has to know where cryptography lives. That sounds trivial. It is not. Keys hide in application code, configuration files, build pipelines, TLS terminators, database columns, and the one spreadsheet a contractor maintains. Finding them is a project, and projects have salaries attached.

Industry surveys consistently show most organizations cannot produce a complete cryptographic inventory on demand. The migration forces the issue. Expect security teams to spend a meaningful share of their next two budget cycles on discovery tooling and outside help, and expect consultancies to price accordingly. This is the least glamorous line item and probably the largest single cost for mid-sized firms.

Code Signing and the Trust Chain#

One underrated expense: signing infrastructure. Firmware, software packages, and container images all depend on signature schemes that must migrate. A signing key change ripples through update servers, customer devices, and any partner that verifies signatures. Getting this wrong bricks devices. Getting it right means coordinated rollouts with rollback plans, which means engineering time nobody budgeted for.

What to Watch in the Next 18 Months#

The signals that separate forecast from reality will be specific. Watch for the first major cloud provider to publish post-quantum handshake overhead numbers. Watch for hardware security module vendors to ship firmware support and for the pricing of those modules to move. Watch procurement language: when governments start requiring post-quantum readiness in contracts, supplier spending follows within a quarter or two.

And watch the insurance market. Cyber insurers have spent years pricing breach risk. If underwriters begin asking about cryptographic inventory the way they ask about backups, that questionnaire will do more to move budgets than any standard.

The migration is not optional and the timeline is not fully under anyone's control. Harvest-now-decrypt-later attacks mean data encrypted today may already be at risk. The bill arrives whether or not the work starts. The only variable is whether it arrives as planned spending or as an emergency.

More articles