Korea Privacy Regulator Reviews Naver AI Tab Search Agent
South Korea’s privacy regulator approved the result of a prior adequacy review for Naver’s AI Tab search agent. AI Tab will provide conversational search answers and may use user activity, age group, gender and interests for personalization. The case offers an early Korean reference point for privacy controls around consumer AI agents.
South Korea's Personal Information Protection Commission has approved the result of a prior adequacy review for Naver's AI Tab search agent, setting privacy conditions for a personalized AI search service.
According to Artificial Intelligence Newspaper, the commission reviewed Naver's plan at its 10th plenary meeting on May 27.
The prior adequacy review system is designed for new technologies and services where existing interpretations of personal information law may not clearly answer how compliance should work.
What Naver is building
AI Tab is described as a search AI chatbot inside Naver Search.
Instead of only listing web pages, the service summarizes and analyzes information and provides it in a one-to-one chat format.
For more personalized responses, Naver wants to use information such as a user's past service activity, gender, age group and interests.
The source says relevant service data could include search records as well as activity from Naver services such as blogs, cafes and shopping.
That makes the service strategically important and legally sensitive.
Personalization can improve search relevance, but it also raises questions about consent, transparency, data minimization and user control when AI systems draw on broad platform activity.
Why the review matters
The commission's decision is an early Korean example of regulators shaping AI agent services before mass rollout.
Rather than waiting for a violation case after launch, the prior review process allows a company and regulator to agree on suitable compliance measures during development.
For Naver, the review can reduce legal uncertainty around AI search personalization.
For other Korean internet and AI companies, it offers a reference point for how consumer AI agents may be assessed when they combine conversational interfaces with user data from multiple services.
Wider signal
Search is becoming one of the main battlegrounds for AI agents because users expect answers, summaries and task support rather than lists of links.
Korea's approach shows that the competitive race will not be shaped only by model quality.
Data governance and privacy-by-design choices will also influence which AI agent services can scale.
The case is therefore relevant beyond Naver.
It highlights how major platforms may need to prove that personalization, consent and safety controls are built into AI products before those products become part of everyday consumer search.
