AI Risk Wire

Cases › made things up › LR-2026-0471

Canada immigration AI reviewer error fabricated job duties in rejection

Preliminary Why: Publisher not yet named · subject to change

Status: action reported, no fix

Severity moderate
Reported 25 Mar 2026

No source states: loss or records, developer and model, agent tool, could do, deployer, industry.

Severity moderate: test: data of others touched with a bounded scope stated; the refusal affects one applicant's own permanent residence application

A Canadian immigration refusal letter cited job duties in a permanent residence case, including control circuit wiring and robot panel programming, that the applicant, a health scientist, said she had never held or submitted. The department said the duties did not match her claimed Canadian work experience. A disclaimer disclosed for the first time in refusals that generative AI supported application processing, that generated content was verified by an officer, and that generative AI was not used to make or recommend a decision. The applicant noticed the description had nothing to do with her, and her lawyer said the system had hallucinated it. Critics raised concerns about generative AI in immigration assessment. The department did not answer questions about which tools it used or what officer verification involved. The lawyer requested reconsideration and the file has been reopened.

What happened

1 · How it was caughtThe applicant noticed after receiving the letter that the job description had nothing to do with her.

What was done

Her lawyer requested reconsideration and the file was reopened.

Risks

Read from the sources; not yet reviewed by a person.

Controls

Not yet read for controls. Controls that mitigate the primary risk:

Application

Grounding and fact checks · Human approval for high-impact actions · Output filtering and redaction · People told they are dealing with AI

Evidence

No source states: what the AI was doing, what went wrong, what it reached or did, when it happened, when it was found, how many records were exposed, the direct loss, a count of the other parties affected, any court case, the cause, what set it off.

Show the full account: 1 items from the primary sources

What happened, in order

Not dated in the sources:

  • Applicant's lawyer requested reconsideration and the file was reopened.
Show every field, with its source and grade

What happened

Earliest source we hold
2026-03-25
D inferred by us from the described behaviour, reason shown
Who caught it
the victim
C one secondary source states it
What went wrong
made things up
C one secondary source states it
primary risk
RK14
D inferred by us from the described behaviour, reason shown

Impact

Severity
moderate
D inferred by us from the described behaviour, reason shown
Why this severity
test: data of others touched with a bounded scope stated; the refusal affects one applicant's own permanent residence application
D inferred by us from the described behaviour, reason shown

Notes on gaps

The department disputes that generative AI played any role in the decision, and this is reported as its position only. The year 2026 for the late February refusal is inferred from the article date. Reopening the file is not recorded as a control change.

LR-2026-0471 · version 1 · published 2026-11-01 after automatic checks; not yet checked field by field by a person · JSON