Is this a cyber incident plan?

It overlaps with cyber and data protection, but AI incidents also include wrong outputs, bias challenges, model drift and supplier changes.

When should the board see an AI incident?

When the incident affects people, sensitive data, material decisions, regulatory exposure, service continuity or public trust.

What is the first thing to do in an AI incident?

Contain the use, preserve evidence and name the incident owner. Do not rewrite prompts, delete logs or rely on memory.

Free resilience tool

Generate a first AI incident response playbook before something breaks

AI incidents are rarely just technical. They involve data, people, suppliers, regulators and communications. This generator gives the board a first playbook for triage, containment, evidence and follow-up.

Your sector

Governance maturity

Organisation size (optional)

What AI systems or scenarios worry you most?

Which incident scenarios should the playbook cover? (choose at least one)

Personal or confidential data entered into AIWrong or harmful output acted onBias, discrimination or fairness challengeVendor outage or model changePrompt injection or security issueRegulator, insurer or auditor queryMedia or stakeholder escalationModel drift or accuracy degradation

What data is involved? (optional)

Personal data of staff or customersSpecial-category data (health, beliefs, biometrics)Data about childrenFinancial or payment data

Choose at least one area to continue.

What the board needs to decide

Who can pause the AI system immediately?
Which incidents trigger data protection, legal, communications or board escalation?
What evidence must be preserved before anyone changes the system?

Frequently asked questions

Is this a cyber incident plan?: It overlaps with cyber and data protection, but AI incidents also include wrong outputs, bias challenges, model drift and supplier changes.
When should the board see an AI incident?: When the incident affects people, sensitive data, material decisions, regulatory exposure, service continuity or public trust.
What is the first thing to do in an AI incident?: Contain the use, preserve evidence and name the incident owner. Do not rewrite prompts, delete logs or rely on memory.

Find out where your AI exposure sits.

We'll tell you plainly what's worth doing, what isn't, and what a board or regulator will expect to see. No pitch deck.

Book a 15-minute call See what we've built

No obligation · no pitch.