GDE-04 · Regulation & Transparency · 9 min read
AI Disclosure in Commercial Ads (01.07.2026), C2PA Provenance, and Biometric Consent Guide
In AI-assisted advertising production, legal and ethical boundaries are prerequisites built into the pipeline rather than afterthoughts. This guide details visible disclosure rules under Türkiye commercial advertising regulations, the ban on fabricated testimonials, C2PA content credentials, and written biometric consent gates.
Key Operational Takeaways
- Display a visible disclosure inside the advertisement whenever AI staging or digital characters materially influence consumer economic behavior (applicable advertising rules).
- Never depict an AI replica of a real person endorsing or testing a product they did not experience; an AI label never cures a deceptive testimonial.
- Verify documented written consent covering infrastructure, channels, and duration before uploading identifiable faces or voices.
Practical Scenario: Permitted Editorial Portrait & Product Campaign
Multi-angle portrait references uploaded with a model's written commercial release plus the brand's authentic product photo.
The upload gate verifies the 18+ age requirement and logs the written consent reference. Portrait references remain isolated inside the brand's workspace. The resulting editorial frame preserves C2PA provenance metadata and embeds a visible in-ad AI scene disclosure.
An auditable campaign deliverable compliant with personal likeness rights and 01.07.2026 advertising transparency standards.
1. Türkiye Commercial Advertisement Regulation (Disclosure and misleading advertisement rules)
Amendments to the Regulation on Commercial Advertising and Unfair Commercial Practices anchor AI usage in advertising around two core principles: visible transparency and non-deception. Whenever AI is used in a manner that materially influences consumer economic behavior or depicts a digital character indistinguishable from a human, a clear disclosure must appear visibly within the advertisement itself.
The second critical boundary is the prohibition on fabricated consumer experiences. Depicting a real person's likeness or a synthetic presenter as having personally used, tested, or experienced a product they did not actually use is a deceptive commercial practice. Adding an 'AI-generated' badge does not legalize a fake testimonial or an unsubstantiated performance claim.
- Disclosures must be visibly legible inside the advertisement rather than buried in hidden metadata.
- Portfolio records and client review screens must distinguish real source photography from refined or generated layers.
- Price, performance, and official approval claims must always rest on verifiable advertiser records.
2. Written Face/Voice Consent and the 6 AUP Prohibitions
The face or voice of an identifiable real person implicates biometric and personal data protections. In ADSOMI, server-side gates block unauthorized face or voice uploads from entering production. Uploading requires verifying that the subject is 18 or older and that written authorization covering processing on the selected infrastructure, target distribution channels, and duration is on file.
Impersonating public figures, celebrities, politicians, or news organizations (deepfakes), transferring motion from unlicensed social media videos, and submitting minor imagery into generative flows are strictly prohibited under our Acceptable Use Policy (AUP-01..AUP-06). Supplying a tenant-owned provider key (BYOK) never waives these consent and moderation gates.
3. C2PA Content Credentials and the 7-Day Quarantine Rule
C2PA Content Credentials and mandatory attribution markers created in the pipeline are preserved across previews, crops, and 3:4 2K mastering; ADSOMI offers no provenance-stripping or watermark-removal tools. Furthermore, temporary media outputs generated on external infrastructure are downloaded into ADSOMI private quarantine storage within 7 days so raw provider URLs are never exposed to end users.
- Face and voice uploads require documented written consent covering infrastructure, channels, and duration.
- Ads featuring AI staging or characters that influence buyer decisions must carry visible in-ad disclosures.
- C2PA provenance credentials and mandatory attribution notices must be preserved upon export.
Explore Other Production Guides
Field Guide: Turning Phone-Shot Product and Venue Photos into Studio-Grade Ads
A practical guide to transforming phone-shot product and venue photography into studio scenes and social ads while preserving packaging copy and geometry.
True Vector Brand Layers, Verified Price Typography, and Safe-Zone Architecture in Ads
Learn how to protect brand logos, headlines, and prices from generative distortion using authentic SVG/PNG vector overlays and per-format safe zones.
Mastering Short Product Videos: 3:4 (1440×1920) 2K Master vs. 9:16 Vertical Derivative Framing
Understand the four-phase short product video pipeline from approved studio start frame to 3:4 (1440×1920) 2K Lanczos + CAS mastering and 9:16 derivatives.
Multi-Brand Workspace Isolation and Side-by-Side Source-vs-Output Client Approvals for Agencies
How agencies managing multiple client brands prevent cross-brand asset leakage with Brand Hub isolation and run evidence-based client approval portals.
Creative DNA and Multi-Hook Iteration Playbook: Preventing Ad Fatigue with Structured Variations
Systematically improve campaign performance by iterating opening hooks, offer layers, and aspect ratios without regenerating your approved product scene.

