The same year is being used to answer different questions
Researchers, regulators, vendors, and risk leaders are often talking about different problems.
- A researcher may be estimating the earliest plausible quantum attack.
- A regulator may be setting a date for organizations to complete migration.
- A bank may be asking how long customer data must remain confidential.
- A blockchain network may be estimating how long technical and governance changes will take.
Those answers can produce different timelines without actually contradicting one another.
Forecasts depend heavily on assumptions
There is no single, fixed date when RSA becomes vulnerable. Any estimate depends on logical qubit counts, physical error rates, error-correction overhead, gate speed, system reliability, algorithmic improvement, and the amount of time an attacker can devote to the calculation.
Change the assumptions and the forecast changes. A confident date without the assumptions behind it is not much of a forecast. It is a calendar wearing a lab coat.
Migration time matters more than headline dates
Even if 2035 proves accurate, organizations may still need years to inventory cryptography, replace unsupported systems, update identity and certificate infrastructure, test interoperability, revise contracts, validate controls, and remove vulnerable fallback mechanisms.
The technology may move quickly. Enterprise migration rarely does.
Long-lived data creates risk before the machine arrives
Some data must remain confidential for a decade or more. That includes health records, government information, intellectual property, financial data, identity credentials, and sensitive communications.
If the protection period extends beyond the estimated arrival of a cryptographically relevant quantum computer, the exposure already exists.
The question is not whether the data is safe today. It is whether it will still be safe when an attacker finally has the capability to decrypt it.