What is actually in your house
A typical 2026 smart home has 30 to 50 connected devices. Each one negotiates session keys with at least one other device or the cloud. The cryptography under the hood is one of:
- ECDH (P-256), in Matter, Thread, recent Zigbee, Bluetooth LE.
- RSA-2048 or RSA-4096, in older Wi-Fi camera pairing, some legacy IoT bridges, vendor cloud APIs.
- Pre-shared keys, in older Z-Wave, original Zigbee, some industrial-style installations.
All three break on a CRQC. ECDH falls to Shor's algorithm for elliptic curves. RSA falls to Shor's algorithm for integer factorisation. Pre-shared keys are not breakable by Shor but are breakable by physical access plus reverse-engineering, which has its own track record.
Why this is not a 2035 problem
Two reasons it bites earlier.
First, harvest-now-decrypt-later. Adversaries who can intercept your home traffic today (Wi-Fi sniffing on a shared router, ISP-level capture, compromised mesh node) can record encrypted streams and decrypt them later. Camera feeds, lock-actuation logs, voice commands, all retroactively recoverable.
Second, replay and impersonation against device identity. The cryptographic identity baked into a smart-home device at manufacture time is what proves "this is your lock" to your phone. If that identity scheme is RSA-based and a CRQC arrives, an attacker can forge that identity from a captured public key.
The Matter and Thread story
Matter (the multi-vendor smart-home standard) and Thread (the underlying mesh protocol) both use Elliptic Curve Cryptography for device commissioning and session establishment. The Matter spec at version 1.x is not post-quantum. The Thread spec at 1.4 is not post-quantum. Both rely on Internet-IETF EDHOC-family key exchange, which the IETF LAKE Working Group is actively post-quantuming as PQ-EDHOC.
The expected migration path: Matter and Thread adopt PQ-EDHOC profiles once IETF finalises them, vendor firmware updates roll out over 2 to 4 years, replacement-on-failure cycles complete the transition over the device lifetime. So new gear bought in 2026 will not get post-quantum cryptography for a few years yet, and gear already in your home will probably never get it.
What you can actually do
For consumers there is no immediate panic button. Three sensible moves:
- Treat your home network as harvestable. Use TLS 1.3 with hybrid PQC for any sensitive cloud service. Most password managers and a few VPNs already offer this.
- Prefer locally-controlled smart-home setups (Home Assistant, HomeKit-local, Matter-local) over cloud-anchored ones. Local control means an adversary needs physical access to harvest, which is a much higher bar.
- When buying new gear in 2027 onward, ask vendors specifically about PQC roadmap. The ones that have an answer are the ones that will be supported in 2032.
What vendors should be doing
- Specify PQ-EDHOC support in 2026 to 2027 firmware roadmaps.
- Pilot hybrid ECDH plus Kyber on cloud-API endpoints. Survives either primitive being broken.
- Plan HSM replacement for production signing infrastructure. Existing RSA-sized HSMs do not fit Dilithium-2.
- Ship secure-update channels that are themselves post-quantum-protected, otherwise the migration gets undone by an adversary forging update signatures.
Where this connects
Smart-home gear is the consumer-facing tip of the constrained-IoT iceberg. The same cryptographic constraints that bind a Thread mesh bind an industrial wireless sensor, a smart-meter network, and a precision-agriculture telemetry fleet. The PQ-EDHOC work in IETF LAKE serves all of them. See the IoT PQ-EDHOC topic page for the standing perspective and quantum-resilient IoT energy metering for what a 2026-deployable PQC stack on Arduino-class hardware actually looks like.