AI Assurance Pulse Check - Inspired Assurance

Step 2

For each question, tick the box that best describes your organisation today.

1. Governance & Oversight*

	Not in place	Just starting	Mostly, not consistent	Fully in place & managed
Do you have a clear, approved policy that says how AI can and can't be used in your organisation?
Is there someone clearly responsible for managing AI risk, with a way to escalate concerns?

Please complete the fields

2. Quality & Testing*

	Not in place	Just starting	Mostly, not consistent	Fully in place & managed
Before an AI system goes live, is it properly tested for accuracy, quality and unintended behaviour?
When an AI system or model changes, do you re-test it before trusting it again?

Please complete the fields

3. Security & Resilience*

	Not in place	Just starting	Mostly, not consistent	Fully in place & managed
Have you checked your AI systems for ways they could be tricked, manipulated or misused?
Do you know if staff are using AI tools that haven't been approved or reviewed by anyone (“shadow AI”)?

Please complete the fields

4. Operations & Incident Response*

	Not in place	Just starting	Mostly, not consistent	Fully in place & managed
Do you monitor your AI systems once they're live, to catch problems early?
If an AI system caused a problem tomorrow, is there a clear process for responding to it?

Please complete the fields

5. Architecture & Tooling*

	Not in place	Just starting	Mostly, not consistent	Fully in place & managed
Do you have a consistent, planned way of building and testing AI systems, rather than each team doing its own thing?
Can you trace which version of a model or system is running, and where it came from?

Please complete the fields

6. People & Awareness*

	Not in place	Just starting	Mostly, not consistent	Fully in place & managed
Have your leadership team and staff received any training on AI risks and responsible use?
Does your organisation have a culture where people feel comfortable experimenting with AI, build AI skills, asking questions and sharing what they learn?
Do staff feel safe raising concerns or challenging how AI is being used when something doesn’t seem right?

Please complete the fields

7. Vendors & Procurement*

	Not in place	Just starting	Mostly, not consistent	Fully in place & managed
When you bring in an AI vendor or tool, do you check them out properly first (security, data handling, reliability)?
Are AI-related expectations (data use, security, support) written into your supplier contracts?

Please complete the fields

8. Release & Deployment*

	Not in place	Just starting	Mostly, not consistent	Fully in place & managed
Are checks for AI quality and safety built into your release process, rather than done manually at the end (if at all)?
If something goes wrong after an AI release, can you quickly roll it back?

Please complete the fields

9. Infrastructure & Recovery*

	Not in place	Just starting	Mostly, not consistent	Fully in place & managed
Is the technical infrastructure running your AI systems (compute, data, storage) secure and reliable?
If your AI systems went down, do you have a tested plan to recover them?

Please complete the fields