By Alexander Truskovsky, CISO, EigenQ
Much of the early post-quantum deployment has focused on key establishment: how two systems establish shared secret values, known as cryptographic keys, used to encrypt and decrypt their communications. The urgency comes from harvest-now-decrypt-later attacks. An adversary can collect encrypted traffic today and potentially decrypt it later, once a sufficiently capable quantum computer exists.
This migration is already moving beyond pilots. We’ve seen post-quantum key establishment deployed in browsers and messaging, as well as supported by cloud services. Many of these deployments use hybrid approaches that combine conventional and post-quantum cryptography during the transition.
The following predictions for 2027 are my best guidance on where I think the quantum and security industries are heading.
1. Authentication and digital signatures will receive more attention
In 2027, I expect more attention to shift toward authentication and digital signatures. Protecting the confidentiality of a connection does not automatically make the certificates used to authenticate it quantum resistant. The same applies to device identities, signed software and firmware updates.
Here, the concern is not someone decrypting old traffic. It is whether an attacker could forge an identity or signature, impersonate a trusted system, or deliver malicious software that appears legitimate. Companies should inventory the certificates and signatures their systems rely on and test post-quantum replacements across issuers, signing systems and the devices or applications that verify them, prioritizing long-lived systems that are difficult to upgrade.
2. Long-lived devices will make migration more challenging
Post-quantum certificate issuance and signing capabilities are becoming available. But deploying them requires the complete trust path to work: the issuer, signing system, trust anchor and the device or application that verifies the signature.
That is particularly challenging for industrial equipment and embedded systems expected to remain in service for many years. Some devices can be upgraded or retrofitted. Others may need to be replaced because their trust mechanisms cannot be changed safely.
For long-lived equipment, the post-quantum decision starts when you buy it, not when a quantum computer becomes a threat.
3. Vendors will need to demonstrate crypto-agility
Supporting multiple algorithms is only part of the challenge. Organizations also need to change keys, certificates, policies and software without disrupting operations. I expect more customers to ask for demonstrations of interoperability, performance, recovery and migration across environments where some systems have transitioned and others have not.
Crypto-agility should be tested through an actual transition, not inferred from a list of supported algorithms.
4. CISOs will face pressure to show measurable progress
Boards and customers will increasingly want to know who owns the migration, which critical systems remain exposed, and what is blocking progress. Useful measures include coverage of cryptographic dependencies, supplier readiness, funded migration plans and successful end-to-end tests.
Priorities should reflect both how long information must remain confidential and how long devices, signatures and trust anchors must remain dependable.
A critical question for the board is whether the organization can complete its migration before its exposure becomes unacceptable.
5. AI agents will put machine identity and permissions in focus
As organizations give AI agents more responsibility, they will need distinct identities, limited permissions, protected credentials and revocable access. Device and workload attestation can provide evidence about the environment handling sensitive operations, but it must be combined with authorization and monitoring.
A valid identity does not mean every action is legitimate. Organizations also need safeguards that limit what an authenticated agent can do.
An AI agent needs an identity and clear limits and organizations need to remember that authentication alone does not make actions safe from the AI agents.
Closing prediction
My prediction for 2027 is that organizations will expand their post-quantum programs from protecting communications to protecting the identities, software and devices those communications depend on. Vendors will need to demonstrate that their systems can transition safely, and CISOs will need evidence of progress.
Organizations do not need a precise prediction of Q-Day to begin. They need to understand their dependencies, prioritize systems that are difficult to change, and start testing migrations while there is time to resolve failures. Q-Day is not the day to begin your post-quantum migration—it is the day you discover whether you began early enough.
##






