Start with the task, not the brand.
Describe the problem your team needs to solve and the result you will accept. A tool that writes convincing prose may still be a poor fit for a task that needs accurate calculations, reliable citations, or controlled access.
Review what goes in and where it goes.
For academy exercises, use only public information or fictional examples. Before a practice adopts any tool, its responsible reviewers should examine the vendor’s current data handling, retention, access, and deletion terms. Product settings and agreements can change.
Try a small, reviewable example.
Use the same fictional task across the options you are considering. Note the corrections needed, whether source links work, and whether the output follows the requested format. A successful demonstration is not proof that every future answer will be correct.
Check the full cost of the workflow.
Include time spent preparing input, reviewing output, managing accounts, training staff, and fixing errors. Avoid buying a plan solely because a demo appears fast.
Make the decision explicit.
Record approved uses, prohibited data, the owner, and when the decision will be reviewed. A tool should have a clear place in the team’s process, with a way to stop using it if it creates problems.
Written and reviewed by CheckMyDSO. Updated 2026-10-16.
Get new prompts by email
Free AI prompts for the front desk, Tuesday and Thursday at 7 AM ET.