On 10 December 2025 Korea’s Prime Minister chaired the 7th national policy coordination meeting, releasing measures against false and exaggerated advertising that uses AI to disrupt market order. The measures respond to the spread of AI-generated “fake expert” and “fake celebrity” ads, and set out responses across prevention, blocking and sanctions.

Three stages

First, prevention before circulation — labelling. The measures require that, going forward, when AI-generated material is produced, edited and published, the fact that the photo or video was AI-generated must be labelled. This connects with the labelling requirements for generative AI outputs in the supporting rules of Korea’s AI Framework Act. See our note on the Korea AI Framework Act enforcement decree.

Second, blocking during circulation — faster takedown and emergency procedures. The measures call for strengthened monitoring and faster takedown, including completing review within 24 hours and introducing an emergency process to block harmful ads even before deliberation is complete.

Third, stronger sanctions — punitive damages. For entities that knowingly spread false or fabricated information through telecommunications networks, the measures allow damages of up to five times the losses incurred, and plan to raise fines the following year.

Why “punitive damages” matters

Korea is moving AI fake ad governance from “administrative penalties first” to “aggravated civil liability”. The point of punitive damages is to raise the cost of wrongdoing so that spreading false AI ads is no longer commercially worthwhile.

For Chinese companies going global, this deserves particular attention: advertising and content compliance in Korea had often been seen as “leniently enforced”. The introduction of punitive damages changes the risk baseline — once a party is found to have knowingly spread falsehoods, the damages can far exceed expectations.

Four points for companies going global

First, check AI provenance in advertising and marketing material. Where marketing teams use AI tools to generate images, video or copy, check whether they contain fabricated expert figures, celebrity likenesses or exaggerated claims, and label AI-generated material as required. See our note on labelling enforcement.

Second, build labelling capability to the strictest standard. Where a product serves China, the EU and Korea, the AI content labelling baseline should be the union of the three, with configuration room for differences. See our note on the EU AI Act transparency rules.

Third, bring “knowledge” risk into contracts. When working with Korean marketing agencies or MCNs, include warranties on the legality and truthfulness of material, and allocate liability for false material provided by the counterparty, to avoid being found to have “known”.

Fourth, build a rapid response mechanism. The 24-hour review and emergency blocking mean platforms and advertisers must complete verification and takedown quickly. Companies should prepare content checklists and takedown processes in advance. See our multi-jurisdiction service.

Relationship to Korea’s overall regulatory pace

Korea advanced AI-related rules intensively in late 2025: the November pre-announcement of the AI Framework Act enforcement decree focused on labelling and responsible parties, while the December fake ads measures focus on enforcement in advertising. Together they show Korea moving AI governance from framework legislation to scenario-based enforcement.

For companies going global, Korea’s importance is often underestimated. Bringing Korea into the Asian compliance map, with a separate labelling and advertising compliance plan, is a worthwhile investment.

This article is based on Korea’s measures against false and exaggerated advertising using AI and public reporting. It is for general reference only and does not constitute legal advice.