Guide
The AI Transparency & Disclosure Framework, Explained
On August 18, 2026, IAB published AI Transparency & Disclosure Framework V2, an update to the standard it first released in January 2026 for when and how AI involvement should be disclosed across advertising and marketing. If you run or are considering conversational advertising — ads placed inside a live AI conversation rather than on a static page — this is the closest thing the industry has to a shared rulebook right now, and it's worth understanding exactly what it does and doesn't cover.
- Published
- Sept 2026
- Reading time
- 3 min
What Is the AI Transparency & Disclosure Framework?
The AI Transparency & Disclosure Framework takes a risk-based, materiality-driven approach: instead of requiring a label on every piece of AI-touched content, it asks organizations to disclose AI involvement specifically when it materially affects authenticity, identity, or representation. IAB built V2 to translate that principle into practical protocols — concrete guidance for when and how to disclose, rather than a single blanket rule.
The stated goal is to avoid two failure modes at once: under-disclosing, where consumers could be misled about what's AI-generated or AI-influenced, and over-labeling, where disclosures pile up on low-risk content until they stop meaning anything.
What's Inside the AI Transparency & Disclosure Framework V2
The framework provides practical guidance across several content types:
- AI-generated and AI-assisted text, imagery, video, and audio
- Synthetic voices and digital twins
- AI-powered consumer interactions — the broadest category, and the one most relevant to conversational advertising
- Common disclosure principles and use-case triggers, meant to support consistent decision-making across brands, agencies, publishers, and technology partners
Read together, it's a shared vocabulary and decision framework — not a certification, and not enforced by IAB itself.
Why the AI Transparency & Disclosure Framework Matters Right Now
V2 didn't arrive in a vacuum. It landed the same window as a wave of new law: New York's synthetic performer disclosure law took effect in June 2026, California's SB 942 took effect August 2, 2026, and Article 50 of the EU AI Act — which sets its own AI transparency obligations — took effect the same day. Regulators in the US, EU, and Asia have moved from proposing AI-disclosure rules to actually enforcing them over the past year, and IAB's update is the industry's attempt to get ahead of a patchwork of state and regional law with one consistent standard.
IAB is running a workshop on the framework on October 28, 2026 for further industry input — a sign this is still an active, evolving standard rather than a finished spec.
What the AI Transparency & Disclosure Framework Doesn't Cover Yet
The framework's "AI-powered consumer interactions" category is broad enough to touch conversational advertising, but it stops short of a dedicated standard for the specific case that matters most here: a sponsored placement served inside a live AI conversation, sitting next to the AI's own answer in real time. That's a meaningfully different disclosure problem than labeling a synthetic image or an AI-written product description — the ad and the AI's organic response are competing for the same moment of attention, and the framework doesn't yet spell out how that pairing should be labeled.
| Covered by Framework V2 | Not Yet Covered |
|---|---|
| AI-generated/assisted text, imagery, video, audio, synthetic voices, digital twins | A dedicated disclosure spec for ads served inside a live AI conversation, alongside the AI's own answer |
| "AI-powered consumer interactions" as a broad category | How disclosure should specifically work for conversational/chat-based ad placements |
| Risk-based, materiality-driven decision principles | An enforcement mechanism — this is industry guidance, not law |
How Thrad Applies This to Conversational Advertising Today
Thrad doesn't wait for that gap to close. Every placement is labeled as sponsored, and ads are withheld entirely from sensitive conversations by default — the same materiality-driven instinct behind IAB's framework, applied specifically to the ad-inside-a-conversation case the framework hasn't formalized yet. See how we approach safety in AI advertising for the full practice.
As the framework evolves — particularly after the October 28 workshop — expect more specificity on exactly this point. Until then, brands and publishers running conversational advertising campaigns are better served by disclosing conservatively now than waiting for a mandate.
