[Seungso-ui Seolgae] Gwangjang Wins "Naver Crawling Dispute"...Unraveling the Secret Code Behind Illegal Data Collection

A bold bet on technical sophistication...collaboration with engineers
Tracking listings for months and exhaustively analyzing the source code
Proving repeated bulk collection

"Even if information is publicly available on a website, automatically collecting it in bulk and using it without permission constitutes an infringement of database (DB) rights. This ruling is highly significant in that it has put the brakes on indiscriminate attempts at 'free riding' in the age of artificial intelligence (AI)."

Kwak Jaewoo, an attorney in the Intellectual Property (IP) & Tech Group at Gwangjang Law Firm (39th class of the Judicial Research and Training Institute), explained the significance of Naver’s recent victory in a DB infringement lawsuit it filed against Darwin Property, the operator of the real estate information platform "Darwin Brokerage." The Patent Court acknowledged that Darwin Property had repeatedly and systematically crawled (automatically collected) and used Naver Real Estate data, and ordered the company to pay 80 million won in damages and to delete the data.

Kwak Jaewoo, an attorney at law firm Gwangjang, is being interviewed by The Asia Business Daily at the firm's office on Namdaemun-ro in Jung-gu, Seoul. Photo by Cho Yongjun

◆Uncovering hidden evidence in the "source code"

The key issues in this case were, first, whether real estate listings that were made publicly accessible to anyone constituted a protected "DB," and second, whether the act of crawling those listings was unlawful. Darwin Property argued, pointing to Naver’s status as a large platform, that "Naver did not directly collect the data, so it has no rights, and the process involves only 'temporary copies' created during linking, which should be exempt from liability."

Gwangjang Law Firm chose to compete on "technical sophistication." Led by attorney Kwak, who graduated from the Department of Industrial Engineering at Seoul National University, the team worked with Naver engineers to thoroughly analyze the "source code" of the Darwin Brokerage website. They discovered, among other things, that the source of listings was explicitly labeled in the source code as "source: Naver," and that Naver’s unique complex identification numbers had been copied as they were, and used these findings to persuade the court.

Attorney Kwak explained, "Due to time constraints, we could not present millions of listings, so we tracked and collected representative listings from across the country over several months," adding, "We demonstrated in real time that when a typo was corrected on Naver, the same correction appeared on Darwin Property’s site after a certain delay, thereby proving the repetitiveness and systematic nature of the bulk collection."

◆"Even public data can be infringed if it has 'substantial volume and quality'"

The Patent Court accepted Gwangjang’s arguments and found that Darwin Property had "illegally copied Naver’s DB, into which a substantial volume and quality of investment had been made, thereby infringing its economic interests." Naver’s efforts to introduce a verified-listing system to filter out false listings, and to invest considerable costs and manpower to update and verify listing information, were recognized as conferring rights as a DB maker.

This case follows in the line of the 2017 Saramin and 2022 Yanolja "HereO" rulings, and is expected to serve as an important benchmark in future disputes over data crawling and AI training. In particular, it is seen as clarifying the scope of platform companies’ rights at a time when scraping data for AI training is becoming increasingly common.

Attorney Kwak stressed, "Rather than indiscriminately scraping other people’s data in search of an easy path, it is time to pursue legitimate license agreements or develop new technical methods that do not involve copying," adding, "I hope this ruling will serve as a minimum bulwark for protecting domestic content and IP."

This content was produced with the assistance of AI translation services.