TKOResearch
Menu
All tools

AI Risk Briefing Builder

Put AI risk, ownership and the next decision on the same page.

Build a board briefing around a defined system and a decision someone can act on. Record the assumptions behind each rating, describe the business impact, and assign an owner before exporting the report.

This is an authored reporting aid, not a risk model or certification. For broader risk-management context, see NIST AI RMF 1.0 (2023). The matrix and display bands here are TKOResearch conventions, not a NIST scoring method.

Build the briefing

Report content stays in this page's memory until you explicitly export or print it. This tool has no account, report API, autosave or local storage. Reloading or leaving this page loses the current report. JSON files contain the report content unless redaction is on.

Define the assessment period and what ratings 1 through 5 mean in your assumptions. Leave a rating unknown when you cannot justify it. Up to 20 risks; 1000 characters per text field.

Report inputs

Risk references follow register order and are renumbered after removal.

Preview and export

When on, every entered text field and date is removed from the preview, JSON, SVG and printed report, including organization, title, system boundary, context, assumptions, risk statements, impact descriptions, controls, names, owners, decisions and rationale. Ratings, confidence, control status, counts and numbered references remain. Inputs above remain editable. This switch does not analyze or anonymize arbitrary prose.

JSON preserves the editable report. SVG is a presentation visual with shortened risk labels. Print includes the full visible report; choose Save as PDF in your browser and review page breaks. Browser headers and footers can be disabled in the print dialog.

AI risk briefing

AI risk briefing

Not supplied · Not supplied

System boundary
Not supplied
Business context
Not supplied
Assumptions and rating basis
Not supplied

Risk summary

0 risks · 0 rated · 0 unknown

Control status: unknown: 0 · planned: 0 · in-progress: 0 · implemented: 0 · verified: 0

Scores multiply ordinal likelihood (1–5) by impact (1–5). They are user judgments, not measured probabilities or predicted financial loss. Unknown inputs have no score. Control status does not change a rating. Scores describe the current assessment; no residual score is inferred.

5 × 5 risk matrix

Likelihood increases upward. Impact increases to the right. Each reference identifies a risk in the register below.

Rows: likelihood · Columns: impact
Likelihood ↑ / Impact →12345
5—————
4—————
3—————
2—————
1—————

Unknown / not plotted: None.

Cell shading groups products 1–5, 6–14 and 15–25 for orientation only. These bands are this tool's display convention, not an acceptance threshold or a standard.

Decision and owner register

No risks added.

Prepared with the TKOResearch AI Risk Briefing Builder. Ratings and statements are supplied by the report author. This report does not establish control effectiveness, certification or a delivered assessment.