The operations manager of a charity with six staff and no HR department asks Claude: "You are a world-class HR consultant. Write a hybrid working policy for my organisation." The draft is well written but refers to an HR business partner, a staff council and an employee benefits portal, none of which the charity has. What is the most likely cause?
- AThe prompt gave Claude a role but no facts about the charity, so it assumed the set-up of a typical large employer. Correct
- BThe role was not specific enough, and a more detailed expert persona would have produced a policy that fits.
- CPolicy writing is a legal task that Claude cannot do, so the charity needs an employment solicitor instead.
- DThe chat used a fast, lower-cost model that defaults to generic templates for any policy-drafting request.
Why A is correct: Claude had no information about the charity's size, structure or existing arrangements, so it filled the gaps with common assumptions. Adding that context, such as six staff, no HR team and who approves requests, would let it write a policy that fits.
Why B is wrong: Since the prompt opened with a role, refining the role feels like the natural fix. A role shapes tone and approach, but no persona, however detailed, tells Claude how many staff the charity has or what support it lacks.
Why C is wrong: Caution about employment matters is sensible, and a specialist may review the final policy. The draft failed because of missing facts about the charity, not because Claude cannot draft a policy, so abandoning Claude over-corrects.
Why D is wrong: Generic output can look like a model limitation. Nothing in the stem points to the model, and any model would have to guess the charity's set-up when the prompt does not describe it.