An attacker can record encrypted communications without alerting the victim and store them until a quantum computer powerful enough to break the keys protecting them becomes available. The US National Institute of Standards and Technology (NIST) describes this scenario in its final report on the post-quantum transition, published on 27 October 2026. It sees an urgent threat: adversaries can collect encrypted data now and decrypt it once quantum technology matures. The risk therefore depends on two dates: the unknown date when such a machine will exist, and the date until which the data must remain secret, which each organisation can determine. For medical records, trade secrets or classified information, that second date is often several decades away.

Collection that goes unnoticed

Most encrypted communications begin by establishing a session key, commonly through an elliptic-curve key exchange. A sufficiently powerful quantum computer could break that exchange, recover the key and read what it protected. The encryption of the content itself, using AES-256 for example, is not susceptible to this attack, but it only protects the data as long as the key established at the start remains secret. An attacker who records the key exchange along with the rest of the session retains what they need to decrypt it once such a machine exists [1] [2].

Passive interception can happen in transit without leaving a trace on the target’s systems. In twelve years of tracking advanced threat groups, I have seen espionage campaigns collect information for years before being discovered. Collection of encrypted traffic outside the victim’s systems might never be discovered. Crisis management usually starts with detection, then focuses on containment and restoring services. Here, detection may never come. If the data is decrypted one day, its confidentiality cannot be restored, although the organisation can still limit some of the consequences.

Deadlines already set

Nobody knows when such a machine will exist. Germany’s guidance of 12 October 2026 cites a study commissioned by the Federal Office for Information Security (BSI) and updated in 2025 [3]. It suggests that a quantum computer capable of breaking current cryptography is likely to be feasible by 2040 at the latest, although nothing yet shows that one exists. Authorities are moving ahead because cryptographic transitions are slow. NIST notes that full integration of a new cryptographic standard has historically taken ten to twenty years [1].

NIST plans to deprecate quantum-vulnerable public-key algorithms after 2030 and disallow them in its standards after 2035. A federal memorandum it cites (OMB M-26-15) requires the most sensitive federal systems to use post-quantum key exchange by the end of 2030 [1]. The coordinated roadmap of EU member states, which is non-binding, sets the same deadline for high-risk use cases [4]. It recommends starting the first steps by 31 December 2026.

Which data to protect first

An intercepted password loses its value once it is changed. A medical record, a manufacturing formula or biometric data can retain theirs for years. The European roadmap treats a use case as high risk if a breach of confidentiality ten or more years from now would still cause significant harm [4]. In such cases, it says attackers have a strong incentive to capture the data now. Business owners need to make this classification, since the CISO cannot determine alone how long a contract, process or personnel record must remain secret.

The classification then needs to be matched against an inventory of the flows that carry those data beyond the organisation’s boundaries, along with the cryptography that protects them. The UK National Cyber Security Centre (NCSC) sets 2028 as the deadline for completing this discovery phase [5]. It recommends recording what data the organisation holds, its expected lifetime and its value to an adversary. Priority flows include communications between sites and between the organisation and its cloud services, partners and suppliers. If AI agents exchange information with outside services or agents, those exchanges also belong in the inventory [2]. For each flow, the organisation needs to know who chooses the cryptography: the organisation itself, its hosting provider or its network operator.

Decisions organisations can make now

Hybrid mechanisms are already available. They combine a classical algorithm such as X25519 with a post-quantum algorithm such as ML-KEM, which NIST standardised in August 2024 [6]. When properly designed, a hybrid construction preserves session confidentiality as long as one of the two algorithms holds. BSI recommends these hybrid approaches [3], and NIST identifies TLS 1.3 as a practical near-term route to adoption [1]. Rollout also means checking equipment and supplier commitments. The NCSC recommends communicating requirements to suppliers by 2028 [5]. A “post-quantum” label alone is insufficient. Organisations should ask which independent evaluations cover the products and configurations on offer.

Germany’s guidance also examines post-quantum cryptography as part of the state of the art [3]. It notes that section 30 of the German BSI Act requires important and particularly important entities to apply the state of the art in cryptography in line with their risk management. Article 32 of the GDPR likewise treats the state of the art, costs and risks as criteria for choosing security measures, including encryption [7]. In my view, an organisation that keeps classical key exchange for personal data requiring long-term protection should at least document that choice and reassess it as post-quantum solutions become more widely available.

A company can start before buying anything: identify what needs to remain confidential for ten years or more, and determine what key exchange protects it in transit.

Sources