How to take on AI responsibilities at work (and get credit for it)
By TechlyUpUpdated 2 min readEmployees and team leads
Quick answer
Start with a small, low-risk pilot on a task your manager cares about. Agree the scope, the data rules, and how success will be judged before you start. Run it for a few weeks, document time spent, quality checks, and problems honestly, and present the method so others can reuse it. That record is what turns side experimentation into recognised responsibility.
Choose a pilot your manager will value
Pick a task that is frequent, visible, and low-risk: weekly reports, meeting summaries, first-draft responses to common requests. Avoid anything involving sensitive personal data or final decisions about people.
Write a one-page proposal
A short written proposal shows maturity and protects you if something goes wrong.
Task: weekly customer feedback summary. Current effort: about 3 hours, one person (estimate). Proposal: AI drafts themes from exported comments; I verify against a sample of 30. Data: comments only, no names or contact details; approved tool only. Duration: 4 weeks. Success: summary accepted by team lead without major edits; time recorded weekly.
Measure honestly
Record time and quality each week, including weeks where the AI draft needed heavy rework. Honest numbers build trust and help decide whether to continue.
Turn results into responsibility
Share a short write-up and offer to help one colleague adopt it. Ask for the work to be reflected in your goals or role description. Visible, documented improvements are hard to overlook in reviews.
Mistakes that undermine AI pilots at work
Good ideas fail for avoidable reasons.
- Starting without telling your manager, then presenting results that raise data concerns.
- Choosing a task no one cares about, so success goes unnoticed.
- Overstating early results and losing credibility when they don't hold.
- Keeping the method in your head, so the improvement disappears if you're away.
Presenting your results
Keep the presentation to one page or five minutes: the task, the method, the checks, the measured results with any caveats, and a recommendation. Include one example of a problem the AI created and how your check caught it — this reassures managers that the method is safe.
End with a specific ask: permission to extend the pilot, time to train a colleague, or inclusion of the work in your goals. A clear ask turns a good result into recognised responsibility.
Try it yourself
Draft the one-page proposal for a pilot in your team using the template above.
Frequently asked questions
What if my company has no AI policy?
Ask before using any tool with work data. Proposing guardrails in your pilot can help your team create a policy.
How long should a pilot run?
Long enough to see several cycles of the task — often a few weeks for weekly work.
What if the pilot fails?
A documented failure with clear reasons is still useful. It shows judgement and saves the team from repeating it.
Want a suggested next step for your situation?
Share a few details and someone from TechlyUp will get back to you. No automated sequences.
Sources and further reading
Examples are authored practice material, not measured learner outcomes. Tool behavior can change. Found an error? Contact TechlyUp with the page URL and correction.