Cadence
An assessment is a snapshot of a legal position at a moment. Positions move, deployments change, and neither event announces itself — so the question is not whether your assessments go stale but how you find out that they have.
Four ways a filed assessment stops being true
Most programmes watch for the first and are surprised by the other three.
The provision moved
Amended, renumbered, superseded, or commenced on a date that has now passed. This is the one everybody expects, and the one a citation-level watch can actually catch: you need to know which of your documents cited the thing that changed.
The deployment moved
The same tool is now used for a second purpose, or rolled out to a fourth state, or pointed at a new data source. Nothing in the law changed and the assessment is nonetheless about a system that no longer exists in that form.
The vendor moved
A rewritten privacy policy, a new sub-processor, a changed retention default. Your conclusions about the vendor were reasonable when written and are now unverified.
The interpretation moved
The text is identical and the regulator has published guidance that reads it differently. This is the hardest one to detect and the one most likely to be raised back at you.
What we watch today, stated plainly. Citation drift on linked assessments, vendor privacy-policy drift, and official-template publication. Sub-processor lists, data-processing agreements and training-policy pages are on the roadmap and are not watched yet. Deployment change and interpretive shift are not automatable and we do not claim them.
Setting a cadence that survives contact
An annual review assumes risk arrives on a schedule. It does not — it arrives when somebody in a department starts using an existing tool for something new.
A review prompted by a citation moving or a deployment being edited is a review with a reason. One prompted by the calendar is a review somebody will defer.
Where the determination is deterministic, re-running it against the current corpus is cheaper and more reliable than a human re-reading last year’s document.
A review that concluded nothing had changed is a finding, and it is dated. Without the record it is indistinguishable from a review that never happened.
A deliberate design decision
When the corpus improves, documents you already received stay exactly as issued. A document that rewrote itself after you circulated it would be worse than a stale one — you would have no idea which version your board approved.
Instead you are told which documents depended on what moved, and you decide whether to regenerate.
Material change is also what sizes the fee for watching — one detector, reading the definition Colorado and California already wrote, serving both the record and the price.
Find out what already aged
Run the same deployment through a current determination and compare. The delta is the backlog you did not know you had.