Board AI Risk Workbench

Two jobs for the board. Score how well it can govern four residual risks of agentic AI today, then begin the register that turns each scenario into a transfer outcome and a board treatment.

Board / company

As of

Board current-state reading

Row averages: the mean of the four questions for one risk. Column averages: one question across all four risks.

Four residual risks · score each on four questions

Pick the sentence that is true today, not the one that is planned. Hover any level for its anchor.

Each risk engages three or four board action lenses. Open Evidence to weigh to see the regulations to consider, the board's structural response, and what management should be doing — the material a board uses to justify a score.

Reading the scale

One anchor per level, per question.

The register

One row per scenario. Click a row to open it. Pass one sets the facts and the outcome; pass two tracks the residual.

Hover a header for what the column means. Materiality and net residual scores feed the Board Risk chart.

No entries yet. Start blank or pull a scenario from the library.

Seven transfer outcomes

Choose one per scenario. The outcome fills the interpretation and the board treatment.

Why now

The market is sorting silent AI into affirmative cover, constrained cover, or exclusion.

Board risk

Materiality across, net residual up. The top-right corner holds the risks that remain large after controls and insurance.

Risks and next steps

Ordered by net residual, then materiality. Suggested steps come from the outcome's board treatment and the division of labor; edit them in the entry.