Two Home Affairs Officials Suspended After AI Hallucinations Entered Official Policy

Two officials submitted policy containing hallucinated information. The submission moved through review. The review layer did not fact-check the content. The policy was approved. Later someone noticed the information was fabricated. The officials were suspended. Adequate has their names and titles.

This is the standard failure mode of review processes designed for volume. The reviewer checks that forms are filled correctly, that citations exist, that tone matches guidelines. The reviewer does not verify the claims themselves. Verification requires expertise, time, and a job description that explicitly authorizes doubt. None of these were present. Suspension feels like accountability.

The policy may or may not be reversed. The officials will be rehired or fired depending on internal politics. New guidelines will be written. They will specify that AI outputs require verification. No one will be assigned to do the verification. The next hallucination will enter policy through an identical gap. Adequate will note the names of the next officials to suspend.