Loading...
Loading...
EU AI Act Article 50. California SB 942. IPTC 2025.1. C2PA. If your studio ships AI-generated work to clients, these apply. Numonic captures the provenance as you create, so disclosure is a setting on export, not a separate workstream.
AI transparency rules are evolving fast, and they are an opportunity, not a burden. Numonic captures every asset’s full provenance the moment it is made, so meeting a new requirement is a change to your export profile, not a scramble. We track the standards as they evolve so you do not have to.
AI transparency requirements are evolving rapidly. Here is how Numonic keeps you compliant.
Disclose when content is AI-generated. Mark deep fakes and synthetic media in machine-readable form.
Auto-embed AI disclosure in metadata. Export with machine-readable AI-origin markers.
Provide provenance documentation for synthetic media on request.
Complete generation history stored. IPTC 2025.1 injection for verifiable provenance.
Standardized fields for AI-generation disclosure in image metadata.
Export with IPTC AI fields populated: DigitalSourceType, AISystemUsed, AIGenerationInfo.
Cryptographically signed content credentials proving origin and edits.
Detect existing C2PA manifests on upload. Display provenance badges for signed assets.
Detailed guidance for providers and deployers on meeting Article 50 transparency requirements.
Disclosure and labeling for synthetic media becomes mandatory for businesses serving California residents.
Machine-readable metadata and detection required. Penalties up to 3% of global annual turnover.
Every export sorts metadata into three categories, so you can prove AI origin without revealing your prompts or workflows.
Removed on export when you share externally.
Kept so the file stays technically usable.
Injected so disclosure travels with the file.
News organizations, magazines, and digital publishers using AI for images, illustration, or enhanced photography.
Agencies and studios creating AI-assisted campaigns for clients with global audiences.
In-house teams using AI to scale content production across markets.
Companies building AI generation tools that need to help their customers meet disclosure requirements.
It requires providers of AI systems that generate synthetic content to mark outputs in a machine-readable, detectable format, and deployers of deepfakes to disclose that content is AI-generated or manipulated. The transparency provisions take effect in August 2026, with penalties up to 3% of global annual turnover.
The AI Transparency Act requires disclosure of AI-generated content and labeling of synthetic media. It takes effect in January 2026 and applies to businesses operating in California that create, publish, or distribute AI-generated content.
It is the latest Photo Metadata Standard, adding AI-disclosure fields including AISystemUsed (engine/model name), AISystemVersionUsed, AIPromptInformation (a generic prompt indicator that does not expose the prompt), AIPromptWriterName, and DigitalSourceType (e.g. trainedAlgorithmicMedia).
C2PA is an open standard for cryptographically signed, tamper-evident content provenance. Numonic detects C2PA signatures on imported assets from providers like Adobe, Microsoft, and the BBC, and displays verification badges in the library. Full validation and re-signing are planned for Q3 2026.
Export sorts metadata into three categories: A (strippable) removes prompts, seeds, and GPS; B (preserved) keeps technical fields like dimensions and color profile; C (compliance) injects IPTC 2025.1 AI fields and C2PA. You disclose AI origin at the right level without revealing trade secrets.
Images (PNG and JPEG with full IPTC/XMP) and video (MP4 and MOV via XMP UUID injection, WebM and MKV via native Matroska tags). C2PA detection covers PNG and JPEG today, with video detection planned for Q2 2026.
EU AI Act Article 50 enforces in August 2026. Start building your AI transparency infrastructure now, so it is already there when regulators, clients, or insurers ask.
Start freeNo credit card to begin