Quantum Shelf Life
THE TEST THAT DOES NOT DEPEND ON PREDICTING THE FUTURE: if your data must stay secret for longer than it will take you to protect it, the exposure already exists.
Quantum Shelf Life
Enter how long one class of data must stay confidential and how long your migration will take. Get your protection margin in years, the latest year you can still start, and whether the exposure already exists.
Protection margin = years until a capable quantum computer, minus the years your data must stay confidential, minus the years your migration will take
A negative margin is different from an ordinary risk. Traffic harvested today cannot be recalled. Chapter 2 puts it plainly: there is no way to retroactively protect data that has already been captured. A negative margin cannot be closed by future effort, only limited.
Nothing you enter is transmitted or stored. Everything is computed in your browser.
The data
One class at a time. Chapter 2 is explicit that different data types carry completely different quantum risk.
Where does that confidentiality period come from?
The single most challenged number in this calculation. A figure taken from an actual obligation survives review. An estimate does not.
Your migration
How long it will take, and when you will actually begin. Both matter, and the second one more than people expect.
Where does the migration duration come from?
A duration nobody has tested is the assumption that most often turns out optimistic, because it usually excludes the vendor who has not committed to a date.
How complete is your cryptographic inventory for this data class?
You cannot migrate cryptography you have not found, and anything undiscovered extends the duration above.
What may already be gone
The part of this that cannot be fixed by any future migration.
Has this data class been in a known exfiltration or breach?
Chapter 2: data taken before migration cannot be encrypted again. If it is already gone, the damage is done and the investment is to prevent further damage.
The planning date
Used for the headline only. The result is tested across the whole plausible range regardless of what you choose.
Who approved the planning date?
Choosing a quantum date is a governance decision. A date nobody has approved is a date nobody will defend when it is challenged.
Ready.
The ask
The decision, the amount, the owner and what cannot be recovered, in the order a board wants them. Everything below is the evidence.
Protection margin
0
years
Within tolerance
The answer that does not depend on the date
The same test run across every plausible quantum date. This is the panel that answers the objection that nobody can predict the timeline.
Swipe the table sideways to see every column.
| Quantum date | Years available | Years needed | Margin | Latest start | Position |
|---|
The latest year you can still start
Exact, not estimated. Work back from the planning date through the migration and the confidentiality period.
0
latest defensible start year at your planning date
0
years of slack, or years already late
What cannot be recovered
The part of the exposure that no future migration closes.
0
years of traffic that may already be held
0
year this class stops being sensitive
What a CFO will attack, and by how much
The two inputs this tool already flags as most challenged, each solved exactly for the point at which the position clears.
0
longest confidentiality period that still clears, against the figure you entered
0
longest migration that still clears, against the figure you entered
The price of another year
Chapter 2’s timeline paradox: a fixed deadline forces action, an uncertain one permits delay. This is what each further year of delay costs.
Swipe the table sideways to see every column.
| Delay | Start | Margin | Position |
|---|
Do your own answers agree with each other?
Individually plausible inputs can still contradict one another. These are the contradictions a reviewer finds first, and they are checked across fields rather than one at a time.
Confidence in this reading
Graded on the four inputs that decide whether the number survives challenge, not on how comfortable the answer is.
Evidenced
The reading a sceptic would take
Earliest plausible quantum date, and every unproven duration extended. This is the number to present.
0
challenged protection margin, in years
0
challenged latest start year
The conversation Chapter 2 records
Marcus Webb had this exchange with his CFO and it went badly. These are the three reframes the chapter gives, populated with your own figures.
You want me to approve twenty million dollars over five years to protect against computers that don’t exist yet, based on a timeline that nobody can predict, for a threat that might never materialize?
When Agents Rule, Chapter 2The threat has already materialised
The investment is not wasted in the best case
The requirement is arriving anyway
Board paper
Written to be pasted into a board or risk committee paper without editing. Replace the bracketed fields before circulating.
Turn the margin into a programme
One data class is a finding. These instruments take it across the estate.
Find what protects it
The Cryptographic Census workbook inventories every cryptographic dependency, scores quantum exposure using this same margin arithmetic, and assigns each asset to a migration wave you can defend. It carries the shelf life test for every major data class in one sheet.
Get the workbookFix the vendor constraint
Your migration duration is capped by your slowest critical vendor. The Vendor Post Quantum Readiness Questionnaire extracts dated commitments rather than roadmaps, with a weighted scoring rubric and the red flags that override the score.
Get the questionnaireUnderstand the countdown
Chapter 2 sets out the harvest now, decrypt later threat and the timeline paradox. Chapter 6 covers the architecture. Chapter 11 builds the quantum economics a CFO will fund, and Chapter 13 works the five year migration programme.
Get the bookThe deadline was yesterday
Every day of delay accumulates debt that compounds faster than you can repay it.
About this calculator. The shelf life test is set out in When Agents Rule by Steven Oppenheim, Chapter 2, and applied across the estate in the Cryptographic Census workbook in the companion toolkit. This tool is a modelling instrument, not an audit, a prediction or a benchmark, and it reflects the figures you enter. The confidentiality periods offered as presets are illustrative starting points drawn from the risk profile in Chapter 2 and are not legal advice on retention. Quantum planning dates are a governance judgement, not a forecast: published assessments vary widely and the tool deliberately reports across a range rather than asserting one date. Nothing you enter leaves your browser: there is no account, no transmission and no storage. This is not legal, technical, cryptographic or financial advice, and it is not a substitute for qualified counsel or your own cryptographic assessment.