How We Test Authenticity
When It Matters
- →Fake or manipulated intimate images, including image-based abuse
- →Photos or video produced as evidence in family law and protection order matters
- →Disputed CCTV, dashcam and body-worn footage
- →Impersonation scams using a fake video of an executive or family member
- →Defamation and harassment involving fabricated images
- →Insurance claims supported by edited or recycled photos
Accuracy and Limits
No single test proves an image is real or fake, and automated deepfake detectors give probabilities, not answers. We combine several independent methods, report the confidence of each finding, and state what the evidence cannot establish. A heavily compressed copy from social media holds far less information than the original file, so we always ask for the earliest version available.
If you are dealing with intimate images shared without consent, you can also report it to the eSafety Commissioner, who can have material removed.
Related Guides
Common Questions
Can you tell if a photo or video is fake?+
Often. We examine the original file, its metadata, compression history and visual consistency. Some findings are conclusive; others can only be reported as consistent or inconsistent with editing.
Do you need the original file?+
It helps a great deal. Files forwarded through messaging apps or social media lose metadata. Send the earliest copy you have and tell us where it came from.
Can AI detection tools settle the question?+
No. Automated detectors produce false positives and false negatives. We use them only alongside file and source analysis, and explain their limits in the report.