Planning
- Do we have senior sponsorship and buy-in?
- Which teams/experts need to be involved in the planning process?
- Do we have clearly defined objectives to be achieved using AI?
- Do we know what the benefits and risks are associated with the proposed AI?
- Do we have the budget to design, develop and implement new AI tools?
- How long will it take to deliver the project?
- Could this process or objective be achieved without expensive intelligent AI tools?
- Do we understand the legal, privacy, compliance, security impacts of building and/or deploying AI solutions?
IT and people infrastructure
- Can we access the data needed to train the AI model?
- What prompts do we need to develop to arrive at an accurate outcome?
- Do we have an AI governance policy related to data, algorithms, intellectual property, ethics etc.?
- Do we have the right skillset to design, develop and implement AI tools?
- Do we have the right skills and IT systems to deploy the AI solution?
- Do we know what questions to ask suppliers when evaluating different solutions?
- Which AI methods should be used in this case given pros and cons of each technique?
Implementation
- Do we have a process for dealing with AI outcome errors?
- How do we train staff to use the new models?
- What controls do we need to put in place to manage the AI risks?
- What feedback did we receive during the testing phase, which should be considered now?
- What implementation documentation supports the maintenance of the AI tool in the future?