The audit-trail vendor for "we did not let AI train on this." Dedicated team. Regulator-grade evidence. Bespoke per-customer policy that survives EU AI Act proceedings.
Forensics is incident-shaped — we run diagnostics when something goes wrong. Defensive is program-shaped — we run the continuous evidence pipeline and sit inside your compliance function. This is the tier that gets named on your SOC2 System Description and your EU AI Act Article 53 filing.
You already have a 40-person GRC team. You need a specialist vendor for the AI-exposure slice, with evidence that plugs into your existing control framework.
PHI adjacency matters. HIPAA, GDPR Art. 9 (special category), and FDA software-as-a-medical-device contexts all require provable training-exclusion for sensitive content.
Material non-public information, classified adjacencies, and pre-disclosure windows. Selective AI visibility is becoming baseline hygiene, not a differentiator.
You license your IP. You negotiate with foundation-model providers. You need evidence of the exclusion windows to enforce the licenses.
Your customers ask: "Is my data visible to AI training crawlers?" The Defensive tier is the answer you can show them in writing.
National archives, judicial systems, and regulators with their own public-facing content and their own reasons to care who trains on it.
Tell us about the shape of what you need. We'll scope an engagement and come back with a specific proposal within five business days. Everything we discuss is under NDA — we'll send one if you don't have a preferred form.
Defensive pricing lands six-figures annually for most scopes we've modeled. We're a two-person shop augmented by our AI council, so we can genuinely only take on a handful of Defensive customers per quarter without breaking our own SLAs. If what you need is Forensics-shaped, start there — you can always grow. If you need the full program, tell us now so we can plan capacity honestly.