Explore our new Cybersecurity Terms Glossary, explained by our experts.
Cybersecurity Terms Glossary
Harvest now, decrypt later describes an adversary recording encrypted traffic today in order to decrypt it once quantum computers can break the key exchange that protected it.

What Is Harvest Now, Decrypt Later?
Most cryptographic risk is about what an attacker can do now. This one is about what they can do later. An adversary with storage and patience captures encrypted traffic today — VPN sessions, TLS connections, database replication, backup transfers — and keeps it until a cryptographically relevant quantum computer can break the key exchange that protected it. The session was secure when it happened and is readable afterwards.
The phrase is also written as store now, decrypt later, and it is the reason post-quantum migration is treated as urgent rather than as a problem for the 2030s. Data encrypted today with X25519 or RSA is already exposed if it must stay confidential beyond the arrival of that machine.
Capturing traffic requires a vantage point and disk space, both of which a state-level actor has in abundance, and neither of which leaves much trace. Decryption is deferred to a future in which Shor’s algorithm, running on a large fault-tolerant quantum computer, solves the integer factorisation and discrete logarithm problems that RSA, Diffie-Hellman, and elliptic curves depend on.
Symmetric cryptography is far less affected. Grover’s algorithm offers a square-root speed-up against a brute-force search, so AES-256 keeps a comfortable margin and AES-128 becomes merely adequate. The exposure is concentrated in the key exchange, which is precisely what a recorded handshake contains.
Michele Mosca framed the deadline as a simple comparison. If x is how long your data must stay confidential, y is how long migration will take your organisation, and z is how long until a quantum computer can break your algorithms, then you have a problem whenever x + y > z.
The value of z is genuinely unknown; expert surveys place meaningful probability on the 2030s. What organisations control is x and y. A bank holding mortgage records for decades and needing four years to migrate is already past the line on any plausible estimate. A retailer whose session data is worthless in a fortnight is not.
| Data | Confidentiality lifetime | Priority |
|---|---|---|
| Health records, genetic data | Lifetime of the individual | Highest |
| Government, defence, diplomatic material | Decades, often statutory | Highest |
| Intellectual property, research, source code | Years to decades | High |
| Legal and financial records, long-term contracts | 7 years and beyond | High |
| Identity documents and long-lived credentials | Until reissued | Medium to high |
| Session cookies, short-lived API traffic | Hours | Low |
Note that signatures are a different case. A signature verified today cannot be forged retrospectively by a future quantum computer, so signing algorithms face a deadline driven by the lifetime of the trust anchor rather than by recorded traffic. Confidentiality is the urgent half.
NCSC guidance published in 2025 sets out three milestones for UK organisations: identify and plan by 2028, complete the highest-priority migrations by 2031, and finish by 2035. NIST standardised ML-KEM, ML-DSA, and SLH-DSA in August 2024, giving the algorithms to migrate to, and has signalled that RSA and elliptic curve cryptography should be deprecated around 2030 and disallowed by 2035.
Treating harvest now, decrypt later as speculative misreads the cost structure. The recording is happening whether or not the decryption ever becomes possible, and the mitigation for new sessions is already deployable. Where the data you move today still matters in ten years, migration planning belongs on this year’s roadmap, and we help organisations scope that work within our cyber risk analysis services.
Bulk collection of internet traffic by state agencies is well documented, and storage is cheap enough that targeted recording is within reach of many actors. No one can confirm which specific flows are being kept. The sensible planning assumption is that high-value traffic crossing untrusted networks may be retained, and to act accordingly for data with a long confidentiality life.
Mostly in transit, because that is what an outsider can capture. Data at rest encrypted with AES-256 is not meaningfully threatened by quantum computing. The exception is stored material protected by a public key scheme, such as encrypted email archives or files encrypted to an RSA key, which share the same exposure as recorded traffic.
Enable hybrid post-quantum key exchange on anything that supports it, starting with web servers, load balancers, VPN concentrators, and inter-site links. Move symmetric encryption to AES-256. Then map where long-lived confidential data crosses networks, and upgrade those paths first. All three steps are available now and require no new standards.
Less urgently for the recorded-traffic problem, because a signature cannot be forged retrospectively for a document that was already verified. The pressure on signatures comes from longevity instead: roots of trust, firmware signing keys, and long-term document signatures must still be valid when quantum computers exist, so hardware with fixed algorithms needs the longest lead time.
You May Be Interested In These
Cybersecurity Terms
17 November 2026
What Is the Difference Between Controller and Processor?
Read MoreWe use a few cookies that are needed to make this website work, and they are always on. We would also like to set analytics cookies, to see which pages are useful; functional cookies, so that our live chat works; and advertising cookies, to measure our campaigns. We will not set any of those until you say yes, and you can change your mind at any time from the Cookie Settings link in the footer.
Choose the cookies you are happy for us to use. Strictly necessary cookies are always on; everything else is off until you turn it on.
These cookies keep the website working: they hold your security checks, your form submissions and the choice you make here. The site cannot run without them, so they are always on.
These let our live chat open and remember your conversation, and let the newsletter sign-up work. Switch them off and the chat window will not appear.
These tell us which pages are read and how people find us, so that we can improve the site. The reports are about groups of visitors, never about you personally.
These let us measure our advertising and show you our adverts on other websites, such as LinkedIn, Google and Meta. Switch them off and you will still see adverts, but they will be less relevant.