Advertising in the AI era is governed by a three-tier labeling framework: high risk (prohibited), medium risk (labeling required), and low risk (no labeling). Medium risk covers AI that could materially influence consumer decisions, such as virtual influencers, AI-synthesized visuals, and personalized messaging, with sponsored content requiring clear labeling.

High-risk ads include illegal activities, rights infringements, false or exaggerated claims, or use of deepfakes or unauthorized likenesses, which are disallowed even if an AI label is present.

ASCI has long been a self-regulatory body for advertising since 1985, collaborating with government and launching its ASCI Academy in 2023 to train stakeholders on responsible advertising.

Enforcement remains a challenge as ASCI issues guidelines and relies on brands to comply and on informed consumers to act as a de facto enforcement mechanism.

Agency inputs are incorporated into coverage, highlighting industry involvement in shaping the rules.

AI in Indian advertising is here to stay, with brands required to follow ASCI’s clarified ground rules for synthetic content.

Responsibility for compliance lies with advertisers, agencies, and publishers, regardless of whether content is generated via prompts or other AI-assisted processes.

ASCI has issued draft guidelines to ensure responsible labeling of AI-generated content, with stakeholder consultation open through mid-June 2026; the framework aligns with the IT Intermediary Guidelines Amendment Rules 2026 to balance transparency with user experience.

The proposed guidelines aim to increase transparency around synthetically generated ads while avoiding label fatigue and aligning with the Information Technology Amendment Rules of 2026.

The overarching stance is that AI can be used in advertising so long as content remains truthful and disclosures are made when necessary.

The guidelines exempt AI-assisted administrative and text uses, including generating ad copy and accessibility descriptions, so long as no false records are created.

Labels are mandated only when AI-generated content changes the perceived truth of the ad; otherwise, labeling is not required to prevent fatigue.