One scan for private info and hidden data across PDFs, images, audio, and video — with one-click fixes: black out, mute, or remove. Nothing leaves your browser.
One scan for private info and hidden data across PDFs, images, audio, and video — with one-click fixes: black out, mute, or remove. Nothing leaves your browser.
GPS coordinates, author and editor names, camera and lens details, editing software and dates, embedded files, and hidden JavaScript in PDFs — the same inspection GhostTrace runs before a scrub.
PII in whatever text the file carries: a PDF's text layer, text inside images (OCR), a spoken transcript (on-device Whisper), or document text — checked for numbers, emails, SSNs, cards, IBANs, IPs, and optionally names.
Drop any file — PDF, image, audio, video, or document. It opens in the studio with Scan for private info pre-selected. Nothing is uploaded.
Review the severity-graded report: high (SSNs, cards, IBANs, GPS), medium (emails, phones, IPs, names), low (dates, organizations). Previews are masked; click to reveal on your own screen.
Fix it in one click — black out findings in a PDF or image (burned in; check the boxes visually), remove an image's hidden data, or mute the speech segments. Or use the per-item tools for granular control.
Two tiers in one scan. Metadata: GPS coordinates, author names, camera details, editing software and dates, embedded files, and hidden scripts in PDFs. Content: PII in whatever text the file has — checked for phone numbers, emails, SSNs, credit cards, IBANs, IP addresses, and optionally names.
No. Every engine — PDF parsing, OCR, speech transcription, name detection, redaction — runs on your device. You can watch the Network tab while a scan runs.
PDFs get the destructive GhostRedact pipeline — each detected finding gets a box and every page is rasterized, so no selectable text remains; look over the result to confirm every sensitive area is covered. Images get a metadata scrub only — text visible in the picture is not removed. Audio and video get the speech segment containing each finding muted. Your original is never modified in place.
No — it is a client-side review aid. Automatic detection catches structured data reliably and names with an on-device model (auto-loaded on capable devices), but no scanner replaces human review for legal obligations. Treat the X-ray as the fastest first pass, not the final word.
Learn more: What hidden data is in a PDF? · Redacting PHI for HIPAA