Perceptual Hash Inversion Attacks on Image-Based Sexual Abuse Removal Tools
Perceptual hashing, crucial for detecting and removing image-based sexual abuse (IBSA) online, faces vulnerabilities from low-budget inversion attacks based on generative artificial intelligence. We advocate implementing secure hash matching in IBSA removal tools to mitigate potentially fatal consequences.
Authors
- Christian Weinert (ORCID: https://orcid.org/0000-0003-4906-6871)
- Maryam Mehrnezhad (ORCID: https://orcid.org/0000-0002-4223-6885)
- Teresa Almeida (ORCID: https://orcid.org/0000-0001-7393-3379)
- Scott Hawkes (ORCID: https://orcid.org/0009-0005-6063-3939)
Institutions
- Royal Holloway University of London (GB)
- Umeå University (SE)
Publication Details
- Journal
- IEEE Security & Privacy
- Published
- 2024-11-21
- DOI
- https://doi.org/10.1109/msec.2024.3485497
- Citations
- 1
- Primary Topic
- Digital Media Forensic Detection
- Type
- article
- Field-Weighted Citation Impact
- 0.25
Funders
- Engineering and Physical Sciences Research Council