Naver has moved to block unauthorised collection of posts on its Cafe community service and to prevent data provided through its official search API from being used for AI training, answer generation or resale. As the value of search and content data rises, it is directly managing the scope of external use.

◆Blocking external sites from filling pages with Cafe posts

Naver recently confirmed cases in which some external sites automatically collected secondhand trading posts and informational content uploaded to Naver Cafe and displayed them in their own services.

When users post sales and purchase listings in a camera trading cafe, external sites take post titles and product information and present them as, for example, a "camera secondhand trading roundup". Clicking a post links to the original Cafe page, but the external operator can fill its site and gain search inflows and traffic without directly gathering users or content.

Naver said in a recent Cafe notice it will ban the mass collection of posts and comments or copying them to external sites without the consent of members and the Cafe. It said the ban also covers automated collection using crawlers, bots or scripts, unofficial access to data and acts that bypass blocking measures.

It also does not allow collected posts to be published on external services or processed and provided again. It said it has asked businesses already using Cafe data to stop collecting it and delete it, and will consider legal action if rights violations are confirmed.

Naver said the notice does not directly target AI training. It said AI crawling of user-generated content has been blocked by default since last year and the notice responds to recently confirmed cases of unauthorised collection by external sites. Separately, Naver is also limiting how data provided through its official search API can be used by external AI.

◆Search API also barred from AI training and answer generation

Revised terms for the Naver Developer Center search API will take effect from the 7th. The new terms stipulate the search API may be used only for displaying Naver search results.

The API is a channel that allows external developers to pull Naver search results by category, including news, web documents, blogs, cafes and Knowledge iN, into their websites or applications.

Under the revised terms, it is prohibited to input data received via the search API into AI or use it for model training, improvement or evaluation. It also restricts exposing that data in AI-generated answers or results. This means it also seeks to block methods that create answers by calling up Naver search results each time a user asks a question, even without directly training an AI model.

The revised terms also prohibit providing or selling search data to third parties, attaching ads to pages displaying search results, or earning separate revenue using the search API. Naver is placing constraints on using its search results as a resource for other services or businesses beyond simply displaying them.

Naver is also reorganising its search API offering system. In July, it halted new applications for the search API, search term trends and Shopping Insight API and began transferring them to Naver Cloud's "Naver API Hub". It excluded shopping, book and specialist material search APIs from the transfer and ended those services. The revised terms, including restrictions on AI use, will also apply to the API Hub from the 20th.

◆Search and content data that determine AI competitiveness

Naver opened its search API in 2006 to allow external developers to link search results to their services. But as generative AI has spread, search results have become a resource used not only as information to display but also for AI answers and model improvement.

Naver also views data and content as central to its own AI competitiveness. Kim Kwang-hyun (김광현), Naver's chief data and content officer, said at a media roundtable in May that the centre of AI platform competition is shifting from model performance to data quality and service competitiveness.

Naver's AI search service, "AI tab", answers user questions and links to purchases and reservations based on data accumulated across integrated search, shopping, Place, blogs and cafes. As search and content data underpin its in-house AI services, this is seen as limiting external AI from using them for answer generation and service development.

Naver has not fully halted external provision of search data. It is keeping the basic purpose of displaying search results in external services, while restricting their use for AI training or separate profit-making businesses.

An industry official said, "As generative AI spreads, search and user content are becoming a platform's core assets, and the scope of data openness is also changing," adding, "Rather than stopping data provision itself, Naver appears to be seeking to directly manage the scope of external businesses using it for AI or separate profit-making projects."