What you get
A scored AI risk assessment across ten weighted domains, with the reasoning that produced the number and a versioned record you can hand to a reviewer.
Risk score: 68/100 · Elevated · Confidence: Moderate Why not higher: human review is in place for every consequential output, and the vendor holds a current SOC 2 Type II. Why not lower: training-data provenance is undocumented, and the fallback path when the model is unavailable has not been tested. Critical advisory condition — Consequential decision influence without adequate oversight: not triggered.
Illustrative extract showing the shape and depth of the output. Not captured customer work, and not a template you receive — your document is written from your own answers.
What is checked, and what is not
Checked automatically
- The score is reproduced by a deterministic rule engine; a model-proposed score outside the engine's band is not adopted.
- A critical advisory condition applies a score floor no matter what the narrative says.
- A second model reviews the completed assessment for meaning, and cannot change the score.
Not checked: your responsibility
- The accuracy of what you told us. Every input is taken at face value and none is independently verified.
- Whether your obligations under a specific law or contract are met. This is a risk rating, not a compliance determination.
- Model behaviour in production, or how the system performs on data it has not seen.
Who it is for
Anyone who has been asked how they govern AI and needs a defensible answer.
With a free account
- The free AI risk scan: 12 questions, scored in your browser, no account.
Paid
- AI Risk Assessment, $299 one-time: the full assessment, run up to 5 times, plus 3 documents.
- Also included in the Readiness Bundle ($899).
Prices in US dollars, one-time, exclusive of tax. Full detail on pricing.