Scenario simulation

AI governance simulation: get an AI assistant approved, launched and under control

You are the executive accountable for getting a generative AI assistant live in a regulated organisation. Make around eight decisions, see what each one costs you, and finish with a governance score and the gaps to close. It takes about ten minutes.

Which organisation are you steering through this?

Answer these three questions if you like, so we can compare results by sector and size. They are optional. We never ask for your name, email or organisation.

How it works

What decisions will I make?

You will make about eight decisions, from what to do about AI already in use through to responding to an incident after launch. Some choices send you down a different path.

  1. Discovery. Staff are already using a public AI tool with customer data.
  2. Ownership. Who is accountable for the assistant's outputs.
  3. Risk tiering. How you classify the use case.
  4. Data and privacy. What the assistant can see, and what you check first.
  5. Vendor assurance. Evidence and contract terms.
  6. Testing and human oversight. What you test before launch.
  7. Go-live. How you launch, and who can stop it.
  8. Day 40. An incident tests everything above.
Scoring

How is my score calculated?

Each choice earns points in one of six areas: accountability, risk classification, data protection, vendor assurance, testing and oversight, and monitoring and response. Your score is the share of available points you earned on the path you took. You also get a time to go-live, because governance that never ships is its own kind of failure.

The scenario is fictional and simplified. It reflects common practice, and it is not legal advice. Your organisation's obligations depend on its sector, regulators and contracts.

Privacy

What happens to my answers?

Your choices are recorded anonymously so we can report on how organisations approach AI governance. We store a random session identifier, your choices, your score, and the optional sector, size and AI use answers above. We do not use cookies for this, and we do not store your name, email address or IP address. Read our privacy policy for detail.

Want to run this with your leadership team?

We use scenarios like this to get executives, risk and technology teams to agree on how AI gets governed in their own organisation.

Talk to us.