New Zealand passport photo checker rejects applicant's open eyes as closed
In December 2016 an online passport photo checker run by New Zealand's Department of Internal Affairs rejected a photo from Richard Lee, a New Zealander of Asian descent, with the generic error "subject eyes are closed" even though his eyes were open. Major news outlets reported the system later accepted a different photo and the department said shadowing and uneven lighting commonly cause such automatic rejections.
An automated image-analysis component misinterpreted shadows and small eye geometry as closed eyes, causing false rejections.
Key facts
- What
- In December 2016 an online passport photo checker run by New Zealand's Department of Internal Affairs rejected a photo from Richard Lee, a New Zealander of Asian descent, with the generic error "subject eyes are closed" even though his eyes were open.
- Incident date
- Dec 7, 2016
- Who
- New Zealand Department of Internal Affairs
- Failure mode
- Hallucination
- AI surface
- Computer Vision
- Severity
- Low
What happened
Richard Lee submitted a passport photo to the Department of Internal Affairs' online photo checker in December 2016 and received an automated rejection stating his eyes were closed despite them being open. He posted the rejection on social media and later obtained a passport after submitting another photo or having new photos taken at a post office. News coverage quotes the department saying the "eyes closed" message is a generic error often caused by shadows or uneven lighting and noting up to 20% of photos submitted online are rejected for various reasons.
What broke inside the model
- 01 · TriggerA user asks for a fact, a citation, or a figure.
- 02 · Model stepThe model writes a fluent, confident answer.
- 03 · Control gapNothing ties the claim back to a real source.
- 04 · FailureA fabricated fact ships as if it were verified.
- 05 · ConsequenceThe false claim reaches a customer, a court, or the public.
Confidence holds, and even spikes, as the claim detaches from any source.
The failure occurred in the automated image-analysis component of the online photo-checker, which misclassified visual features around the eyes, likely shadows, uneven lighting, or small eye geometry, as closed eyes, producing a false negative. This indicates a weakness in the system's feature-detection or pre-processing for varied lighting and diverse facial characteristics.
What it cost
Sources
Cite this entry
https://failureindex.ai/failures/zealand-passport-photo-checker-rejects-applicantAI Failure Index. "New Zealand passport photo checker rejects applicant's open eyes as closed" (FI-0470). Realm Labs. https://failureindex.ai/failures/zealand-passport-photo-checker-rejects-applicant (indexed Jun 10, 2026).Data fields CC-BY 4.0, prose citation permitted. Incident ID FI-0470. Full dataset at /data.
Note from Realm Labs, the Index steward
How Realm fits
- Prism
- OmniGuard
- AI Detection & Response (AIDR)
This entry sits in the index's predictive wing: a system that scores, ranks, perceives, or steers rather than generates. Realm's runtime layer is built for the generative and agentic systems now moving into these same decision seats, where it watches a model's internal state and holds an unsupported claim or an unchecked action before it commits. The control gap on this record, an automated decision that reached people with no runtime check in front of it, is the same gap. The index keeps predictive failures on the record because the pattern carries straight into the systems shipping today.