Last week, Google submit revised proposal meeting almost all of South Korea’s security requirements
JAKARTA – Google has spent 19 years seeking permission from South Korea to export detailed 1:5,000‑scale map data. For just as long, Seoul has refused.
The reason most often cited is national security: South Korea is technically still in a state of war, so high‑precision maps in foreign hands are considered a risk.
However, that narrative has weakened after Google last week submitted a revised proposal that essentially meets almost all of the government’s security requirements.
According to straitstimes.com, the one condition Google continues to reject is the requirement to build a data centre in South Korea.
Prof Yoo Ki‑yoon, former Director of the National Geographic Information Institute, the government body that produces Korea’s base maps, argues that framing the issue as a dispute over regulation or server location misses the core point.
“If it makes economic sense, Google will come, pay taxes, and compete. That is what happened in Japan,” said Yoo, a Professor of Geospatial Engineering at Seoul National University, in an interview with The Korea Herald.
“The reason Google has not done so in Korea is not because it is blocked, but because the returns do not yet justify the costs.”
The costs Google is weighing are not limited to operating map servers. “Physical presence would trigger broader scrutiny.”
Google Korea reported revenue of 386.9 billion won in 2024 with corporate tax of 17.2 billion won.
However, estimates presented to the National Assembly by Prof Jeon Seong‑min of Gachon University suggest Google’s actual revenue from YouTube, the Play Store, and advertising in South Korea is between 4.8 and 11.3 trillion won, based on SEC filings and domestic company reports.
The gap reflects a corporate structure that channels revenue through low‑tax jurisdictions.
[[ Low‑tax jurisdictions mean Google structures its corporate entities so that much of the money generated in Korea is not recorded in Korea, but routed to countries with lower tax rates. ]]
A local data centre, Yoo said, would expose that structure to scrutiny, not only for maps but potentially for all Google services used by Korean consumers.
In the domestic market, Naver Maps holds around 68 per cent share, integrated into a search and commerce ecosystem that Google has never truly penetrated.
“A standalone mapping service does not generate enough revenue to justify the tax obligations triggered by a local data centre,” Yoo said.
All commercial maps in Korea are built on top of the government’s base map, a public infrastructure developed with state funding over decades.
The principle is simple, Yoo said: “If you want to use public infrastructure to make a profit, you must be present locally and pay your share.”
Some argue that tax policy and map‑data export should be separated, with tax addressed through reform and map export assessed independently. Yoo considers this unrealistic in practice.
“If you design a system to impose equivalent costs on Google, it will inevitably extend to all other global IT companies in the country,” he said.
Digital‑service taxes or similar levies would need to be applied broadly, increasing institutional complexity and the risk of trade friction.
“A case‑by‑case approach using existing legal tools is the most realistic alternative.”
Yoo also sharply criticised the economic studies shaping the debate. A preview of research by Prof Jung Jin‑do of the Korea National University of Education estimated that exporting map data could cost Korea up to 197 trillion won over a decade.
By contrast, a 2024 study by the Korea Academic Society of Tourism and Leisure projected an additional 3.4 million tourists per year thanks to improved Google Maps.
Another paper (August 2025) cited a Google blog and claimed cumulative industry benefits of 18.46 trillion won over five years if exports were allowed.
“None of these studies is credible,” Yoo said. The 2025 paper cited by Google was not even published in a reputable journal. “It is more like a report than an academic work.” (DK/LM)
as a preferred source on Google
