Skip to main content

Smartbox.ai Gen 2 is in beta. General availability January 2027. Join the beta

Smartbox.ai

Defensible disclosure

What makes a disclosure defensible.

A disclosure is defensible when you can answer four questions, with evidence, long after it went out.

The four questions

  1. 01 · Discover

    Did you find everything you should have?

    Record where you searched, what you found, and why the search stopped where it did. For subject access, the ICO says you must be able to show why a search is unreasonable or disproportionate.

  2. 02 · Decide

    Was each decision lawful, and who made it?

    Every withholding decision needs a legal basis and a named decision-maker, recorded when the decision is made rather than reconstructed afterwards.

  3. 03 · Act

    Is the withheld information actually gone?

    A box drawn over text can be lifted. Redactions have to be burned into the file, and the released copy checked against what you decided.

  4. 04 · Prove

    Can you prove it years later?

    The record has to be assembled as the work happens, in a form that shows whether it has been altered.

What the rules say

Where AI fits

Detection proposes; people decide. In Smartbox no generative model decides what is withheld, and the same input always gives the same result, so every decision can be explained and reproduced.

Not ready for a demo?

Get the Gen 2 briefing.

A short overview of what Gen 2 does and how it would work for your obligations.

We use cookies to measure how the site is used, and — if you agree — to measure our advertising. You can accept one without the other, and declining is one click. See our cookie policy.