AI & Generative AI
Where Generative AI Can Help at Work
A grounded framework for identifying suitable generative AI tasks and reviewing their outputs.
- Published:
Generative AI can help with tasks involving language, summarisation, classification, and drafting. Its usefulness depends on the task, the information available, and how the output will be checked.
Start with a bounded task
Look for work with a clear input and a result that a person can review. Examples may include summarising meeting notes, drafting routine messages, or helping users locate relevant internal guidance.
Provide appropriate context
Use relevant, approved information and follow the organisation's data-handling rules. Do not assume every internal document is suitable for every AI workflow.
Review the result
Generative AI can produce incomplete or inaccurate answers. Define who reviews outputs, how errors are corrected, and when a task must be handled without AI assistance.
Measure whether it helps
Compare the assisted workflow with the existing process. Consider quality, time, rework, user confidence, and review cost.
Key takeaway: Choose a specific task, keep appropriate human oversight, and measure the outcome before expanding AI's role.