AI Transparency & Compliance

1. Our Commitment to AI Transparency

SevenTwelvez is an autonomous performance creative system. We use artificial intelligence extensively across our OS to compress timelines, predict success, and scale outputs. We believe transparency about AI usage is not just a regulatory requirement — it is a competitive advantage. This page documents exactly how AI is used across our platform, what controls exist, and how we comply with emerging global AI regulation.

2. How AI Is Used on This Platform

Our platform delegates specific, bounded tasks to specialized AI models (predominantly the Gemini family) to augment human labor.

• Creative Quality Control (QC Engine): Gemini-powered video specification verification, identifying improper formats, audio levels, or watermark presence autonomously.
• Creative Brain: An intelligence layer that learns patterns from buyer and creator activity to predict creative performance and CTR.
• Creative Cortex: An evolutionary engine that decomposes ad creatives into specific genome traits, breeding top combinations for viral lift.
• Smart Assignment: An AI-assisted load-balancing algorithm matching creator performance histories to specific incoming brand briefs.
• Declyn AI Assistant: A unified conversational agent powering global platform support and creator semantic discovery.
• Ad Copy Generation: Instantaneous generation of platform-specific headlines, primary text, and hooks for autonomous campaigns.
• Content Localization: AI-powered translation and cultural adaptation of creative assets scaling campaigns globally.
• Fraud Detection: ML-driven anomaly detection analyzing click velocity, duplicate submissions, and UTM payload tampering.
• Insurance Verification: Zero-trust OCR extraction and analysis of commercial ACORD insurance certificates for platform whitelisting.
• Routing Intelligence: Predictive quantitative scoring for optimal asset-to-ad-network deployment.

3. Content Provenance & C2PA Compliance

Every creative asset distributed through the SevenTwelvez ecosystem receives a C2PA 2.0 content credential before distribution. This cryptographic wrapper permanently bonds the asset to its true history. The manifest contains:

• Creator Identity: Platform username, authenticated tier, and credit bureau reliability score.
• Authenticity Profile: Computed percentages of human involvement versus AI generation, alongside an overall originality index.
• Genome Traits: If applicable, the decomposed creative hooks and angles categorized by the Creative Cortex.
• License Scope: Approved platforms, duration limitations, geographic bounds, and exclusivity flags.

Every distributed asset carries a cryptographically signed C2PA 2.0 content credential using COSE Sign1 encoding with ECDSA P-384 and an X.509 certificate — the exact format required by the C2PA 2.0 specification — that can be independently verified by any compliant tool. Public verification for any asset is natively available at /verify.

4. Authenticity Classification System

To provide immediate clarity to both buyers and downstream consumers, our Authenticity Verification Layer programmatically classifies every asset into one of four rigid categories based on custody tracking:

• Fully Human: ≥ 80% human involvement, ≤ 20% AI contribution. Traditional UGC or studio recordings.
• Hybrid Assisted: ≥ 40% human involvement. Core human footage heavily edited, scripted, or stabilized by AI workflows.
• Template Derived: ≥ 70% template similarity index. Scaled programmatic variants.
• System Generated: The default classification for AI-primary content with minimal human baseline interaction.

5. AI Disclosure at Distribution

When a creative asset is pushed programmatically to an ad platform (Meta, TikTok, Google) via our Autonomous Engines, SevenTwelvez forcibly attaches machine-readable disclosure metadata. This includes the platform-appropriate AI label (Human-created / AI-assisted / AI-generated), EU AI Act compliance flags, California SB 942 compliance signals, a C2PA manifest reference, and an exact disclosure timestamp. This metadata is stored immutably in our Ledger systems and is available for external regulatory audits upon request.

6. Regulatory Compliance Posture

EU AI Act (Article 50)

Article 50 of the EU AI Act demands explicit disclosure when content is artificially generated or heavily manipulated. SevenTwelvez satisfies this strictly through mandatory C2PA manifests (acting as machine-readable marking) and our distribution metadata (human-readable labeling requirements). Under the definitions of the Act, SevenTwelvez operates responsibly as a "deployer" of general-purpose AI systems.

California SB 942 (AI Transparency Act)

California SB 942 mandates that AI platforms provide robust tools for end-users to identify AI-generated content footprints. We exceed this requirement via our unrestricted public verification endpoint at /verify, the enforcement of C2PA manifests across all exported content, and the native visibility of Authenticity Classifications directly inside buyer execution dashboards.

FTC Guidelines on AI & Advertising

All AI-assisted content scaled through our distribution modules integrates with our Legal Gatekeeper to ensure compliance with the FTC's endorsement guides, including forced verification of appropriate paid disclosure markers (#ad, #sponsored) verified aggressively by our Pre-flight Creative Scanner prior to tokenized delivery.

7. Human Oversight (Three-Layer Governance)

SevenTwelvez enforces a strict "Three-Layer Governance" control system to ensure safety. The pipeline acts as follows:

• Layer A (AI Proposer): Observes metrics and generates typed recommendations. Never executes independently.
• Layer B (Guardrail Engine): Validates actions deterministically against hardcoded spend limits, risk profiles, and legal policies.
• Layer C (Action Executor): Processes only mechanically approved commands.

No AI system on this platform can make irreversible capital or disciplinary decisions without deterministic, non-AI rule validation. "Red Zone" escalations strictly demand manual human intervention. All actions are logged immutably.

8. Data Practices for AI

Your privacy is integral. Our embedded AI models are structurally isolated; they are not trained broadly on generic customer payloads. The localized Creative Brain intelligence profiles are user-specific and rigorously firewalled. Live calls dispatched to external endpoints (e.g., Google Gemini APIs) operate strictly under Zero Data Retention mandates, ensuring your prompt data is never used to train upstream foundational models. Every AI-generated outcome within SevenTwelvez is cataloged permanently for audit integrity.

9. Contact Information

For direct questions regarding our specific AI transparency policies or machine learning implementations, contact our Trust & Security team at ai-transparency@zolvaritech.com.

For broader regulatory inquiries or compliance verifications, please reach out to legal@zolvaritech.com.