How to Spot AI-Generated Images in Hiring
Test ability, not images. Fabricated portfolios and profile photos are real problems, but an image detector is a weak gatekeeper: it produces false positives and says nothing about whether a candidate can do the job. Use image checks only to decide what to verify, protect candidate data while you do it, and settle the question with a live interview or a paid test task.
Where AI images show up
The visible cases are portfolio pieces, illustrations in writing samples, and profile photos. The costlier case is a candidate who cannot reproduce the work they showed. That gap is closed by a skills interview, not by scanning pixels, so do not let an image check become the whole process.
Handle candidate data carefully
Candidate photos, IDs, and portfolios are personal data. Before you upload anything to a checker, confirm that it does not store, log, or reuse uploads. Our free AI image checker analyzes images in memory and never stores them, but the same is not true of every tool. For formal identity checks, use an approved verification provider and follow your jurisdiction’s rules, not a general image detector.
A workflow that holds up
- Check consistency. Do the portfolio images match the candidate’s stated history, style, and other work? Inconsistency is a better clue than any score.
- Check provenance. Some images carry Content Credentials; inspect them at contentcredentials.org/verify.
- Check the source. A reverse image search can show whether a portfolio piece was lifted from someone else.
- Verify the skill. A short live interview or a small paid task tests ability directly and fairly.
If you use a detector, use it correctly
Score an image only as one signal, and never as an automatic rejection. A high score means the image shares traits with generated images, and edits, filters, and compression can push real photos into that range. When something looks off, ask the candidate to talk you through the work. That conversation is more accurate than the tool, and it is fairer. We explain the failure modes in Do AI Image Detectors Actually Work?
Common questions
Should I reject a candidate because a detector flags their photo?
No. A detector returns a probability and produces false positives, especially on edited or compressed photos. Use a score to decide what to verify, and confirm with the candidate before it affects a decision.
Can AI detection verify a candidate identity?
No. Image detection checks the picture, not the person. Identity verification needs a proper process: a live interview, and for roles that require it, a documented ID and right-to-work check.
Is it legal to run candidate images through a detector?
It depends on your jurisdiction and how the data is handled. Candidate photos and documents are personal data, so avoid tools that store or reuse uploads, and check your policy before scanning anything. For formal identity checks, use an approved provider, not a general detector.
Where do AI images actually show up in hiring?
Most often in portfolios, writing samples with illustrations, and profile photos. Fabricated work samples are the bigger risk, and they are usually caught by a skills interview or a paid test task rather than an image scan.
A checker that does not keep your files
AI Detector scores an image and deletes it immediately after processing.
