VestigeForensics

— Report anatomy

A complete forensic report

Every field below ships in the JSON and in the web view. The values here are representative; your report carries your exhibit’s.

Analyze your own, free

Forensic analysis report

VF-9E41C7A2

AI-flagged · strict ≤1% FPR
2026-07-12 14:32 UTCpipeline 2026.07.4-productioncalibration full_v4

Sample · simulated result

Sample document with representative values. The detector scores are illustrative; the thresholds and detection-power figures come from calibration run full_v4.

Exhibit

Filename
exhibit_2216-042.jpg
Type
image/jpeg
Dimensions
1280 × 960
Size
280.9 KB
sha-2569e41c7a2d05b8ff31c26e94a7d10b5c8e2f6a03d914b7ce50a8d21f6b3e49c77

Calibrated decision

0.941

Primary detector score. 0 = real, 1 = AI.

0.00.216 · ≤5% FPR0.725 · ≤1% FPR1.0

The primary score sits above the strict threshold of 0.725. Run full_v4 set that threshold to flag no more than 1% of real photographs from any source. Under Social chain conditions the detector catches 43% of AI images at this standard.

Thresholds are calibrated for the Social chain condition, never a fixed 0.5.

Primary detector · drives the verdict

Community Forensics — VF fine-tune

CVPR 2025 base · ViT-S/16 · 384px · pinned checkpoint

0.941

Supporting signals · corroboration only, never decide

Community Forensics (stock)

CVPR 2025 · ViT-S/16 · 384px

0.212

SDXL Detector

Community (HF) · Swin transformer classifier

0.716

A low supporting score does not weaken the verdict. Stock models are known to lag on 2024+ generators.

Robustness · detection power by condition

estimated conditionSocial chain

Screenshot of a WhatsApp image, which is what actually walks into a lab

Clean
64%
JPEG q70
61%
JPEG q50
62%
Resize 50%
61%
WhatsApp
55%
Screenshot
60%
Social chain
55%

Share of AI images caught at ≤5% false positives, run full_v4.

honesty gap 0.504

Headline ranking quality (AUC 0.933) minus the calibrated detection rate at your chosen standard (43%). It measures how much benchmark performance does not survive the evidentiary threshold, and we print it on every report.

Limits · read before acting on this report

  1. 01A flag is a calibrated statistical signal, not proof that an image was generated by AI. Check provenance and context before you act on it.
  2. 02The encoding fingerprint points to screenshot-style re-encoding. Sensor-level signals cannot be recovered from an image in that state.
  3. 03We expect the stock model to lag on 2024+ generators, and at 0.212 it does. Closing that gap is why the fine-tuned primary drives the verdict here. Supporting signals corroborate; they never decide.
  4. 04At the strict ≤1% false-positive standard the calibrated detection rate is 43%, so the absence of a flag would have been weak evidence that the image is authentic.
  5. 05Generators released after the calibration run may not be represented, and a score on a generator we have never seen can be arbitrarily wrong.
  6. 06This verdict applies to the exact bytes hashed above. Re-export or edit the file and you have a different exhibit.

Chain of custody

image sha-2569e41c7a2d05b8ff31c26e94a7d10b5c8e2f6a03d914b7ce50a8d21f6b3e49c77
report sha-2564c8d1a96f0b3e7255d9c02aa8f16de433b7a90c1e5f2481706bd3c5a92e8f014
report VF-9E41C7A2created 2026-07-12 14:32 UTCpipeline 2026.07.4-production