AI Transformation Quiz: What To Do Next

LET’S TAKE A LOOK

Find out where you’re exposed before something goes wrong.
5 questions. 1 minute. This tells you if you’re compliant or sitting on a legal or commercial time bomb.

Question 1: Do you know which AI tools your staff are actually using and how?

A. Yes. We have an approved tools list. We’ve verified that training features are OFF for everything on it. We know what country the data is processed in. B. Mostly. We trust staff not to put sensitive data into AI tools, or conduct unapproved work. C. Sort of. We have a list of approved tools but haven’t checked their privacy or GDPR settings ourselves. D. Not really. We’ll deal with it if something goes wrong. We’re unlikely to get singled out.

Question 2: If your AI is talking to customers, do they know it’s AI?

A. Yes. It identifies itself immediately, and customers can always reach a human if they need to. B. It’s mentioned in our Terms & Conditions… somewhere. C. No. We want it to feel like talking to a person—that’s the point if training it. D. We use AI for dynamic pricing and conversions, but we don’t tell customers that.

Question 3: Who’s responsible when AI gets it wrong?

A. A competent human reviews every high-stakes output before it goes out. They’re accountable. B. We encourage reviews, but people skip them once they trust the tool to generate reasonable information. C. The vendor. If their AI hallucinates or screws up, that’s on them. We’ll sue them for damages, or claim in their professional indemnity insurance. D. The AI. We assume it’s right unless it obviously breaks. We’ve built one AI system to check the output of another.

Question 4: Have you checked if your AI is biased when it is making decisions on your behalf?

A. Yes. We run bias tests and equality impact assessments for anything high-risk like hiring and finance. B. The vendor says their model is fair. That’s good enough for us. C. We make sure our information’s OK, but we don’t specifically test for discrimination. D. Bias is inevitable with these systems. We do what we can but we’ll probably address it if someone complains.

Question 5: If your AI fails catastrophically tomorrow, what happens?

A. We have a documented incident response plan to recognise the issues, respond clearly, and put a plan in place to stop it happening again. B. We’d fix it fast, but we don’t have a formal protocol written down. C. Our marketing and tech people would handle it, but there’s no specific failure process. D. It’s unlikely something will happen. If it does, incidents get dealt with as they come.

Is your business ready for AI

Find out where you’re exposed TO RISK before the AUTHORITIES DO

Is your business really ready for automation?

MY Readiness Check takes less than a minute

LET’S TAKE A LOOK