Why can't I see where part of an AI draft came from?

Not everything that influences a draft appears as a citation. Some inputs shape how an answer is written rather than supplying evidence about a person's work.

Answer

When a claim in a draft is grounded in a Moment, you can open that Moment and inspect the evidence behind it. Other inputs influence a draft without behaving like Moments and without producing citations:

  • Company Context provides organizational framing.
  • Prompt Instructions shape how the answer is written.
  • Context you added yourself, such as a Note or a File, can inform the draft.
  • 15Five AI memory can influence the interaction and the writing style.

So a draft that reads differently than you expected is not necessarily drawing on evidence you cannot see. It may be reflecting framing or instructions rather than a new performance claim.

Key Rules

  • Moments can be opened and inspected. The writer decides what to keep.
  • Company Context and Prompt Instructions do not produce citations.
  • Something said to the review assistant does not automatically become review evidence. Add it as context if it should ground the review.
  • Every draft can be edited, rewritten, or removed, and you can write an answer yourself instead.

Common Misunderstanding

An uncited sentence is usually framing rather than a hidden factual claim. If a specific factual statement about someone's performance cannot be connected to any context available to you, do not assume an explanation. Raise it with your review admin, who can escalate it with the answer and the available context.

Access and Availability

  • Relevant to review writers completing a self review, manager review, or additional-manager review.
  • Available in the Total Platform and Perform & Goals pricing packages.
  • Applies to manual review cycles. Lifecycle Reviews are excluded from AI-Assisted Reviews and are unaffected by these settings.

Related Articles

Was this article helpful?

Sorry to hear that. Tell us what was missing β†’