• Cybersecurity Glossary
Harvest Now, Decrypt Later: The Quantum Attack already happening
Harvest now, decrypt later is a quantum era attack in which an adversary captures and stores encrypted data today to decrypt it later, once a quantum computer can break the cryptography protecting it. The attack is already happening: the harvesting is now, the decryption comes later. It makes quantum risk a present day concern for long lived data. Urgency is judged with Mosca’s inequality, and the response is to protect the most sensitive, long lived traffic first. This page explains the threat, why it is happening now, Mosca’s inequality, which traffic to prioritize and where Teldat fits.
What harvest now, decrypt later is?
Harvest now, decrypt later, often shortened to HNDL and also called store now, decrypt later, is a quantum era attack strategy in which an adversary captures and stores encrypted data today, intending to decrypt it in the future once a cryptographically relevant quantum computer can break the classical cryptography protecting it.
The phrase describes the two stages of the attack and why its name matters. The harvest happens now: an attacker records and keeps encrypted traffic using ordinary means. The decryption happens later: it waits for quantum hardware powerful enough to break the encryption. Crucially, the attacker only needs a quantum computer to finish, not to start, which is why this is described as the quantum attack that is already happening.
That timing is what turns quantum computing from a distant concern into a present one. Any data captured today that must still be secret when a capable quantum computer exists is effectively exposed the moment it is transmitted. The rest of this page explains why the attack is underway now, how to judge your exposure with Mosca’s inequality, which traffic to protect first, and how to respond.
Why the attack is happening now?
It is tempting to treat quantum risk as something to deal with when quantum computers arrive. Harvest now, decrypt later is the argument that changed that thinking, and these are the reasons it makes the threat a present day one.
Mosca’s inequality: your exposure
To decide how urgent the risk is for your own data, the standard tool is Mosca’s inequality, a simple framework from cryptographer Michele Mosca. It uses three variables, and a single rule, to turn an uncertain future into an actionable decision. The table sets out the three terms.
| Variable | What it means | In your control? |
|---|---|---|
| X — shelf life | Years the data must stay confidential | Set by the data’s nature |
| Y — migration time | Years to migrate to post-quantum cryptography | Yes, the variable you control most |
| Z — threat timeline | Years until a quantum computer can break today’s crypto | No, and it is uncertain |
The rule: if X + Y > Z, you are already exposed. If the time your data must stay secret plus the time your migration takes is greater than the time until a quantum computer arrives, that data will still need protecting after it can be broken. The insight is that you do not need to know Z precisely: long lived data plus a slow migration can already put you past the line. Because Y is the variable you control most, starting the migration sooner is the main lever you have.
Which traffic to protect first?
Not all data is equally exposed to harvest now, decrypt later. The priority is set by confidentiality lifetime: the longer data must stay secret, the more it is endangered. These categories illustrate where to focus first.
How to respond today?
The right response to harvest now, decrypt later is not panic but a deliberate, prioritized start. Because the harvesting is already possible, the goal is to shorten your migration time and protect the most exposed data first.
In practice this means inventorying your data flows and ranking them by how long they must remain confidential, then applying Mosca’s inequality to decide which are already past the line. The flows with the longest shelf life are protected first with quantum resistant cryptography, while shorter lived data can follow. Among the highest value steps is protecting the VPN and IPsec tunnels that carry concentrated, sensitive traffic between sites, since they are a natural target for harvesting.
The protection itself can be deployed today using standardized, quantum resistant techniques for the tunnels, and it is best introduced in a phased way, with crypto agility so algorithms can be updated as post-quantum standards mature. The central message of harvest now, decrypt later is simply that waiting is not neutral: because the clock started when the data was first sent, the sooner the migration begins, the less of today’s sensitive traffic ends up exposed.
Harvest now, decrypt later and Teldat
Defending against harvest now, decrypt later means protecting the traffic it targets, and much sensitive traffic travels through the VPN and IPsec tunnels at the heart of SD-WAN. Teldat approaches this from a European base, with central management to make a prioritized migration workable.
Protect long lived traffic before it is too late: the answer to harvest now, decrypt later is to protect the most sensitive, longest lived traffic first, and much of it flows through IPsec tunnels. Teldat combines SD-WAN with IPsec, central management through CNM and European manufacturing to support a prioritized, quantum resistant migration. Because post quantum standards and product support evolve quickly, the specific quantum resistant capabilities available should always be verified with Teldat for a given product and software version.
Frequently asked questions about harvest now, decrypt later
❯ What is harvest now, decrypt later?
Harvest now, decrypt later, often shortened to HNDL and also called store now, decrypt later, is an attack strategy in which an adversary captures and stores encrypted data today, intending to decrypt it in the future once a cryptographically relevant quantum computer can break the cryptography protecting it. The key point is that the attack is already happening: the harvesting of encrypted traffic takes place now, even though the decryption must wait for quantum hardware that does not yet exist. This turns the quantum threat from a future problem into a present one, because any data captured today that still needs to be secret when a quantum computer arrives will be exposed. It is the main reason organizations are urged to begin migrating to quantum resistant cryptography now rather than waiting.
❯ Why is harvest now, decrypt later a threat if quantum computers do not exist yet?
Because the attacker does not need a quantum computer to start, only to finish. The harvesting step, recording and storing encrypted traffic, can be done today with ordinary means and is cheap, so an adversary can accumulate vast amounts of ciphertext now and simply wait. The decryption step happens later, once a cryptographically relevant quantum computer becomes available. This means the clock on the risk starts the moment data is transmitted, not the moment quantum computers arrive. Any information that must remain confidential for years or decades, and that is being sent today, is therefore potentially already compromised, sitting in an archive waiting to be decrypted. That is why security agencies treat HNDL as the primary justification for acting on quantum risk now.
❯ What is Mosca’s inequality?
Mosca’s inequality is a simple framework, introduced by cryptographer Michele Mosca, for judging how urgent post-quantum migration is for a given set of data. It uses three variables: X, the number of years the data must remain confidential; Y, the number of years it will take to migrate systems to post-quantum cryptography; and Z, the number of years until a cryptographically relevant quantum computer exists. The rule is that if X plus Y is greater than Z, the data is already at risk, because it will still need protection after a quantum computer can break today’s encryption. Its power is that it does not require knowing Z precisely: even with an uncertain quantum timeline, long lived data plus a slow migration can already put an organization past the line. In practice it is used to rank systems so the most exposed data is migrated first.
❯ Which traffic should be protected first?
The priority is data with the longest confidentiality lifetime, because that is what harvest now, decrypt later most endangers. Applying Mosca’s inequality, the traffic to protect first is whatever must stay secret for many years or decades and is being transmitted now: for example health records, financial and legal data, intellectual property and trade secrets, government and defence information, and personal data subject to long retention. Short lived data, such as a value that is worthless once a session ends, is far less exposed, since it will not matter by the time a quantum computer exists. The practical approach is to inventory data flows, rank them by how long they must remain confidential, and apply quantum resistant protection to the long lived, sensitive flows first, typically the VPN and IPsec tunnels that carry concentrated sensitive traffic between sites.
❯ How does Teldat help against harvest now, decrypt later?
Harvest now, decrypt later is addressed by protecting the traffic it targets, and much sensitive traffic travels through VPN and IPsec tunnels, which are central to SD-WAN. Teldat is a European manufacturer of networking and SD-WAN, so this is directly relevant to its portfolio. Teldat SD-WAN builds secure IPsec tunnels between sites, managed centrally through the CNM platform, which is the layer where quantum resistant protections for IKEv2, such as post-quantum preshared keys and hybrid key exchange, apply. Central management supports a phased, priority based migration, letting an organization protect its longest lived, most sensitive tunnels first, exactly as Mosca’s inequality advises. As a European manufacturer under European jurisdiction, Teldat also supports the digital sovereignty goals that matter for the long term confidentiality of data. Because post quantum standards and product support evolve quickly, the specific quantum resistant capabilities available should always be verified with Teldat for a given product and software version.
The clock is already running on your data
Harvest now, decrypt later means today’s sensitive traffic is at risk the moment it is sent. Teldat combines SD-WAN with IPsec, central management through CNM and European manufacturing to protect your longest lived tunnels first, before it is too late.

