Provenance Radar
Scan content

Do you need to disclose this content?

New transparency rules like the EU AI Act's Article 50 don't come with an easy answer. Provenance Radar checks whether a file's own records name a tool you'd recognize — Adobe Firefly, ChatGPT, Midjourney, Canva — plus the underlying credentials, metadata, and watermark signals, so you can make the call yourself.

Scan contentView sample report

Documents and images · More formats coming soon

Provenance Report

student-essay.pdf

JD
Named tool
Adobe Firefly
Metadata
Found
Content Credentials
Found
Watermark signals
No supported signal

Assessment summary

Scoped to the evidence found in this file

Inconclusive

The file's record names Adobe Firefly as a generating tool. The report keeps the underlying credential and metadata details available for review; a named tool alone does not establish authorship.

Named tool · Content Credentials · Metadata · Watermark

Recent checks

View report
  • Adobe FireflyProvenanceNamed in record
  • Document metadataMetadataFound
  • Author metadataMetadataFound
  • Supported watermark scanWatermarkNo detection
  • AI text indicatorsAI signalsElevated
  • File integrityMalware scanPassed
Tool names in plain language. Technical evidence underneath.
Adobe FireflyChatGPTMidjourneyCanvaMetadataContent Credentials
Never a single AI score

See more than an AI score

Content authenticity cannot be reduced to one probability. Provenance Radar brings multiple independent signals together so your team can understand what is verified, what is inferred, and what remains unknown.

Explore the platform

One report that verifies, detects, and preserves the evidence

Start with what a file says about the tools involved, then inspect its credentials, metadata, supported watermark signals, and other indicators. It's built to show evidence, not a verdict.

See named tools clearly

When a file names a recognizable tool such as Adobe Firefly, ChatGPT, Midjourney, or Canva, see that name in plain language alongside the original record.

Verification · Provenance

Check the available evidence

Review a file's credentials, metadata, supported watermark technologies, and AI-generation indicators in one place.

Detection · Signals

Preserve the evidence

Turn findings into understandable reports that can support review, investigation, client workflows, and future audit requirements.

Evidence · Reports

The challenge

AI detection alone doesn't tell the full story

The Binary Detector Problem
Most tools collapse uncertainty into one AI probability, hiding the difference between verified provenance and statistical inference.
The Evidence Gap
Tool names such as Adobe Firefly, ChatGPT, Midjourney, or Canva can be buried in technical records alongside credentials, metadata, watermarks, and other workflow evidence.
The Missing Signal Problem
A missing tool name, watermark, or Content Credential does not prove content is human-created; provenance evidence must be interpreted in context.

Who it's for

One report, three very different jobs

Know what you're publishing, too.

Academic integrity is only half the picture. As institutions take on their own obligations to disclose AI-generated content, get the same evidence-based review for what your school publishes, not just what students submit.

Explore education

Know before your client finds out.

Since August 2026, businesses that deploy AI-generated or AI-manipulated content carry new disclosure duties. Screen assets before delivery so you have evidence relevant to that obligation, not a surprise after the fact.

Check an asset before it ships

Labelling at platform scale.

Since August 2026, platforms deploying AI-generated content face new disclosure requirements. Screen uploads automatically and surface evidence relevant to those obligations, without reviewing every file by hand.

Explore the API

What's actually in every report

A named tool can be useful context, but it is never the whole story. Every scan surfaces independent evidence so your team can judge the full picture — never a rolled-up score.

4

Independent evidence categories in every report — named tools, credentials, metadata, and supported watermark or AI-generation signals

7 days

Retention window before original uploaded files are deleted — the evidence and report stay permanently

2

Export formats per report — a human-readable view and a machine-readable JSON export

From file to evidence in seconds.

  1. 01Content is submitted for a scan
  2. 02Provenance Radar identifies the content type
  3. 03Named tools are surfaced in plain language when the file records them
  4. 04Credentials (C2PA / Content Credentials) and available metadata are examined
  5. 05Supported watermark and AI-generation signals are evaluated
  6. 06Results are normalized into one evidence report
View a sample report

Start free, scale with your evidence needs

Every plan shows recognizable tool names when a file records them, alongside Content Credentials, metadata, and a scoped assessment — never a single AI score. Add history, sharing, and integrations as your needs grow.

Free

For quick provenance checks.

$0/ mo
no card required
  • Limited monthly scans
  • Named-tool and Content Credentials checks
  • Embedded metadata checks
  • Malware scan on every upload
  • Basic evidence summary report
Start scanning

Pro

Most popular

For professionals who need deeper evidence.

$29/ mo
billed monthly
  • Everything in Free
  • More monthly scans
  • Detailed evidence reports
  • Scan history dashboard
  • Shareable public report links
  • Declared watermark & AI-generation signals
  • PDF exportable reports(Coming soon)
  • Batch scanning(Coming soon)
Start Pro

Business / API

For teams, platforms, agencies, and institutions.

Custom
Talk to us about pricing and terms
  • Everything in Pro
  • API access(Coming soon)
  • Team workspaces(Coming soon)
  • High-volume & batch processing(Coming soon)
  • Extended evidence retention(Coming soon)
  • Integrations(Coming soon)
  • Optional white-label(Coming soon)
  • Priority support(Coming soon)
Talk to us

Need something custom?

Talk to us about API access, education programs, agency workflows, and data retention needs.

Talk to us

Developers

One evidence API. Multiple signals.

Coming soon: submit content programmatically and get structured evidence back — including recognizable tool names when a file records them, plus Content Credentials, metadata, and supported watermark and AI-generation signals — normalized into one response your product can use.

Illustrative example only. No API endpoint exists yet.

Get notified when it ships
Request
curl -X POST https://api.provenanceradar.com/v1/scans \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -F "file=@student-essay.pdf"
Response
{
  "id": "scan_8f2c1a90",
  "status": "complete",
  "assessment": "review_required",
  "evidence": {
    "content_credentials": "not_detected",
    "metadata": "found",
    "ai_generation_signals": "elevated",
    "watermark_signals": "unsupported"
  },
  "report_url": "https://provenanceradar.com/report/8f2c1a90"
}

Frequently asked questions

Everything you need to know about how Provenance Radar reads and reports evidence. Can't find an answer? We're one message away.

Talk to our team

Provenance Radar is a content provenance and authenticity evidence platform. Upload a document or image and it analyzes Content Credentials, embedded metadata, supported watermark signals, AI-generation indicators, and other evidence — then brings the findings together in one clear report.

See the full story behind your content

Go beyond a single AI score. See the provenance, signals, and evidence behind the content you review.

Scan contentTalk to us