The flagship of the AI-security suite
Meet EU AI Act Article 50 transparency obligations in one click.
DisclosureGuard is your AI-disclosure copilot — the compliance tool that turns every creative into a filed, defensible disclosure:
- Reads what the file actually declares. C2PA content credentials, XMP, EXIF and frame-level video provenance — never AI-detection guessing.
- Scores it 0–100 against the rules that carry fines. EU AI Act Art. 50, Google Ads synthetic-content rules, China's AI-labeling measures — with exact fix guidance per finding.
- Files the regulator-ready paper trail. One-click label packs and tamper-evident audit reports that prove the disclosure survived publishing.
No credit card. Free at $0/mo: 5 assets/month, basic scan. Pro $49/mo: unlimited assets, all 3 jurisdiction packs (EU · US · China), Drift Monitor, Platform Lens. Full disclosure, the way we'd file it: our scanning is deterministic — declared provenance only, human-verified rules, no AI detection and no guessing.
DisclosureGuard is AI disclosure compliance for the disclosure era: it deterministically parses embedded provenance metadata (C2PA manifests, XMP packets, EXIF fields — never AI-detection guessing) and evaluates each asset against versioned jurisdiction rules for the EU, the US/Google Ads, and China. Free plan: 5 assets per month with a basic scan. Pro $49/month. Team $149/month.
The flagship capability — shown, not claimed
Your label dies on upload — we prove it survived
You attach an AI-disclosure label, upload to Instagram — and the platform recompresses your file, stripping the C2PA manifest, XMP packet and EXIF fields the label lives in. Your asset is compliant at export and non-compliant at publish, and nothing tells you. The Disclosure Drift Monitor re-fetches the live file and diffs it against the publish-time baseline — so you can prove the label survived, and get alerted the moment it doesn't.
Exhibit A · Evidence
🧪 The drift test: 9 out of 9 labeled assets lost their provenance
We ran DisclosureGuard's own drift logic — the same deterministic baseline_for + diff_baseline code that powers the Pro Drift Monitor — on 3 labeled assets pushed through 3 simulated platform-recompression pipelines (Instagram-style, Meta-style, TikTok-style). Result: 9/9 — every asset lost its C2PA manifest and XMP packet; the file bytes changed on all 9.
Lab simulation of platform recompression — real uploads require platform accounts; quality levels approximated from public reports. Full methodology disclosed, results reproducible. Read the full drift-test report →
Run the free drift check on your own image → One free check — no account, no credit card. Go Pro — $49/mo →
Section I — The dossier
What it checks, filed in order
Nine sections, each one a standing part of the compliance file your regulator will ask to see.
C2PA content credentials
Parses JUMBF manifest stores in JPEG, PNG and MP4 — claim generators, assertions, ingredients — the strongest, regulator-preferred provenance form.
Frame-level video scans
Samples frames across the video and parses provenance in each one — because a manifest in the container doesn't guarantee clean frames.
EU · US · China rule packs
Versioned jurisdiction engine: EU AI Act Art. 50 transparency duties, Google Ads "How this ad was made" disclosure, China's explicit + implicit labeling measures.
0–100 compliance score
Per-asset score with grade and severity-ranked findings — every finding carries the exact requirement text and plain-language fix guidance.
One-click label pack
Overlay badges (EN/ES, light/dark), caption copy, Google Ads disclosure text, jurisdiction requirement sheet — export as a zip.
Tamper-evident audit trail
Every check, scan and download is appended to a hash-chained audit log; the PDF report carries the chain head so tampering is detectable.
Scheduled re-scans
Weekly automatic re-scans of a brand's whole asset library, with per-brand compliance trending on the dashboard.
EN + ES labels
Multi-language disclosure text for every market you publish in — English and Spanish out of the box.
Team workspaces
Five seats with owner/admin/member roles, plus API access for brands, scans and findings.
Section II — Sealed exhibits
Three exhibits no competitor files
Researched live against ten competing tools. Each exhibit is a sealed, deterministic capability — no AI guessing anywhere.
Exhibit A
⏱️ Disclosure Drift Monitor
Did your label survive publishing? Re-fetches the live file daily and diffs it against the publish-time baseline — SHA-256 change, C2PA manifest stripped, XMP stripped. Drift lands in your hash-chained audit log, so you can prove compliance at publish time even when Instagram recompression stripped the label.
Pro $49/mo — agencies with publishing pipelines. Get Pro — $49/mo →
Exhibit B
🛫 Ad-Platform Disclosure Pre-flight
Platform Lens simulates, before launch, exactly what Google Ads, Meta and the EU first-exposure rule will detect and display for your asset — verdict per platform plus the exact panel copy and action items. The ad-launch workflow feature.
Pro $49/mo — see exactly what the viewer will see. Get Pro — $49/mo →
Exhibit C
⏰ Regulatory Deadline Clock
A per-brand countdown over every dated enforcement milestone touching your jurisdictions — days remaining or overdue, penalty exposure bands, exact rule citations — plus a fine-exposure calculator (enter turnover → EU Art. 50 exposure = max(€15M, 3% turnover)) and an assets-at-risk rollup.
All plans — on Free too. Fear is the wedge; the clock is the hook.
Filed pricing — no surprises
Compliance costs less than the fine
Start free. Pro pays for itself the first time a label survives publishing — or the first time a regulator asks for the paper trail.
Free
$0
forever
- 5 assets per month
- Basic scan: provenance + EU AI Act Art. 50
- Per-asset 0–100 compliance score
- Regulatory Deadline Clock
Pro
$49 /month
per month · license key
- Everything in Free, plus:
- Unlimited assets
- All 3 jurisdiction packs: EU, US/Google Ads, China
- Disclosure Drift Monitor — daily re-fetch + baseline diff
- Platform Lens pre-flight simulation
- One-click label packs, audit PDFs, weekly re-scans
Team
$149 /month
per month · license key
Just +$100/mo over Pro
- Everything in Pro, plus:
- 5 team seats with roles
- API access
Buy a license key via WhatsApp or email — then redeem it on the billing page.
Section III — Procedure
How the file gets built
Name it, pick where you distribute (EU, US, China) — the right rule packs load automatically.
Upload files or paste image/video URLs in bulk. Declare the basics: is a disclosure label attached? Does it depict a real person? Is it an ad?
Deterministic parsing of C2PA manifests, XMP, EXIF and generator text — plus frame-level scans for video. No guessing, ever.
Severity-ranked findings with fix guidance, one-click label pack, Google Ads disclosure copy, and a regulator-ready audit PDF with a tamper-evident hash.
Section IV — The ledger
Why now — the deadlines are fixed dates
| Where | What changed | What it costs to ignore |
|---|---|---|
| EU | AI Act Art. 50 transparency obligations apply since 2 August 2026: machine-readable markings + visible disclosure of AI-generated content. | Fines up to €15M or 3% of worldwide turnover for Art. 50 breaches. |
| US | Google Ads requires disclosure of synthetic/AI-generated content in ads — misrepresentation policy, enforced at the platform. | Ad disapproval, limited serving, account suspension. |
| China | AI-generated content labeling measures effective 1 September 2025: explicit (visible) + implicit (metadata) labels. | Orders to rectify and fines under applicable rules. |
Figures are from the public regulation and policy texts — verify the current consolidated text before legal reliance. The direction is not in doubt: disclosure is mandatory, and it is auditable.
Section V — The suite
Part of the AI-security suite
DisclosureGuard is the flagship — the reason to buy in. One suite, one login, five tools for the disclosure era:
Vibe-Code Audit
AI-powered security audit for your codebase — find the flaws before attackers do.
Rogue-Agent Defense
Firewall and policy shield for autonomous AI agents — 100% block rate across 55 red-team attack tests.
ChatBreaker
Weekly red-teaming for your AI chatbot — 40+ attacks, proof transcripts, drift tracking.
DisclosureGuard
AI disclosure compliance — provenance reading, jurisdiction checks, label packs, audit trails. You are here.
Section VI — Sworn statements
Frequently asked questions
What are the EU AI Act Article 50 transparency obligations?
Art. 50 transparency obligations (in force 2 August 2026) require providers and deployers of AI systems generating synthetic content to mark it machine-readable and to inform recipients that it is AI-generated. Fines for Art. 50 breaches run up to €15M or 3% of worldwide annual turnover. DisclosureGuard checks your assets against these rules.
Is DisclosureGuard an AI detector?
No — deliberately not. AI detectors guess, and guesses are wrong often enough to be dangerous. DisclosureGuard reads declared, embedded provenance: C2PA manifests, XMP packets, EXIF fields, generator text already in the file. It reports what is present and flags what is missing. That is what regulators and ad platforms actually ask for: demonstrable disclosure, not a probability score.
What does the EU AI Act require, exactly?
Art. 50 transparency obligations (applicable since 2 August 2026) require providers and deployers to mark AI-generated or manipulated content in a machine-readable way and to inform recipients that the content is AI-generated — typically a visible label plus embedded metadata. Fines for Art. 50 breaches run up to €15M or 3% of worldwide annual turnover. Verify the current consolidated regulation text before legal reliance.
What do Google Ads rules require?
Google Ads' misrepresentation and synthetic-content policies require advertisers to disclose AI-generated or synthetically altered content in ads — the "How this ad was made" disclosure. Enforcement is at the platform level: ad disapproval, limited serving, or account suspension. DisclosureGuard generates the disclosure copy for you.
What does China require?
The AI-generated content labeling measures (effective 1 September 2025) require explicit (visible, hard-to-remove) labels plus implicit (metadata) labels on AI-generated content, with watermark guidance for generated images and video.
What is the tamper-evident audit trail?
Every scan, setting change and export is appended to a per-workspace audit log where each event embeds the hash of the previous one. The PDF report prints the chain head hash — re-run the report any time and the hash must match byte-for-byte, or the log was altered.
Does a high score mean I'm legally safe?
No. A high score means the checked requirements passed against the current rules data — it is evidence of diligence, not a legal opinion. Rules change; the jurisdiction engine is versioned (current: v2026.10.1) so every report states exactly which rules ran.
How much does it cost?
Free: 5 assets per month with a basic scan (provenance + EU pack). Pro is $49/month: unlimited assets, all three jurisdiction packs, label packs, audit PDFs, scheduled re-scans, Ads copy generator, EN+ES labels, Drift Monitor and Platform Lens. Team is $149/month: everything in Pro plus 5 seats with roles and API access. Buy a license key via WhatsApp or email and redeem it on the billing page.
What is the AI-security suite?
Five tools, one login: DisclosureGuard (AI disclosure compliance — the flagship), Vibe-Code Audit (code security), Rogue-Agent Defense (agent firewall), BreachHunter (website pentesting) and ChatBreaker (chatbot red-teaming). DisclosureGuard is the wedge: the compliance tool every AI-publishing business needs first.
DisclosureGuard Pro — $49/mo: unlimited assets, all 3 jurisdiction packs, Drift Monitor, Platform Lens.