Salvato in:
| Autori principali: | , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | https://arxiv.org/abs/2605.09203 |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866910205816602624 |
|---|---|
| author | Goonatilake, Yevin Nikhel Ateniese, Giuseppe |
| author_facet | Goonatilake, Yevin Nikhel Ateniese, Giuseppe |
| contents | Watermarks for AI-generated images are meant to support downstream decisions about provenance, manipulation, and trust. In the settings that motivate watermark removal, therefore, success means more than causing the watermark test to fail. A successful remover must also preserve the utility of the image and make the output forensically indistinguishable from clean content, so that defeating the verifier restores deniability rather than merely replacing one detection signal with another. We show that current watermark removal attacks fail this stronger objective. Across six state-of-the-art removers spanning four attack families, independent forensic detectors distinguish removal-processed outputs from clean images at over 98% true-positive rate under a 1% false-positive budget. Thus, current removers often replace the watermark with a different detectable signal. Using UnMarker (IEEE S&P 2025) as a detailed case study, we show that this signal persists under common post-processing, exhibits a characteristic two-regime spectral deformation, and yields a three-way tension among removal success, image quality, and forensic stealth. These results show that existing removal benchmarks are incomplete: they reward verifier evasion and utility preservation while omitting forensic stealth. A workable watermark remover must satisfy all three conditions at once: watermark evasion, utility preservation, and forensic indistinguishability from clean content. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_09203 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Removing the Watermark Is Not Enough: Forensic Stealth in Generative-AI Watermark Removal Goonatilake, Yevin Nikhel Ateniese, Giuseppe Cryptography and Security Watermarks for AI-generated images are meant to support downstream decisions about provenance, manipulation, and trust. In the settings that motivate watermark removal, therefore, success means more than causing the watermark test to fail. A successful remover must also preserve the utility of the image and make the output forensically indistinguishable from clean content, so that defeating the verifier restores deniability rather than merely replacing one detection signal with another. We show that current watermark removal attacks fail this stronger objective. Across six state-of-the-art removers spanning four attack families, independent forensic detectors distinguish removal-processed outputs from clean images at over 98% true-positive rate under a 1% false-positive budget. Thus, current removers often replace the watermark with a different detectable signal. Using UnMarker (IEEE S&P 2025) as a detailed case study, we show that this signal persists under common post-processing, exhibits a characteristic two-regime spectral deformation, and yields a three-way tension among removal success, image quality, and forensic stealth. These results show that existing removal benchmarks are incomplete: they reward verifier evasion and utility preservation while omitting forensic stealth. A workable watermark remover must satisfy all three conditions at once: watermark evasion, utility preservation, and forensic indistinguishability from clean content. |
| title | Removing the Watermark Is Not Enough: Forensic Stealth in Generative-AI Watermark Removal |
| topic | Cryptography and Security |
| url | https://arxiv.org/abs/2605.09203 |