Google “spoofs” NAVER iPhone App as unannounced crawler agent

Confessions S5 – Ep 28 NAVAR What It Seems: Google’s Korean Crawler Mystery

The Discovery – A Strange Crawler Appears

• July 19, 2025: Unusual user agent spotted
• DNS-confirmed Google IP with suspicious identity
• The user agent string breakdown

The Suspicious User Agent String

Mozilla/5.0 (iPhone; CPU iPhone OS 18_0 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/18.4 Mobile/15E148 Safari/605.1 NAVER(inapp; search; 2000; 12.13.0; 16PROMAX) (compatible; Googlebot/2.1; +http://www.google.com/bot.html)

Known NAVER Crawler Agent References

• https://www.useragentstring.com/Yeti1.0_id_17673.php
• https://www.useragentstring.com/Yeti1.0_id_17674.php

Decoding the Deception: How Google Spoofed NAVER

• iPhone 16 Pro Max simulation
• NAVER app masquerade (inapp; search; 2000; 12.13.0)
• Googlebot in plain sight

Why Korea Matters to Google

• 99.5% mobile internet usage
• $5.1 billion search advertising market
• Google’s declining mobile search share vs NAVER dominance

NAVER’s Walled Garden and Google’s Content Access Problem

• Historical Korean website blocks against Google
• NAVER’s walled garden strategy
• “Reluctant to allow content…to be used by other search engines”

The App-First Reality (Key insight)

• 30 million NAVER Cafe users, 23+ million Blog users
• Korean content that exists only in apps, never on web
• Google’s app indexing blind spot in Korea

The Gemini Connection: 7-Month Training Timeline

• Timeline: Korean language support (Dec 2024) → Crawler appears (July 2025)
• Gemini’s need for “subtle differences in meaning, idiomatic expressions, and honorifics”
• The 7-month gap as training period

Google’s Strategic Endgame: Bypassing Korean Content Restrictions

• Bypassing content restrictions through identity deception
• Accessing NAVER app content that’s never been web-indexed
• Feeding Korean language data into Gemini training

Testing Phase Evidence

• Random targets (test sites, German soccer guestbook)
• Technical methodology validation — including how Google processes content signals at crawl time, reveals the deeper mechanics behind these covert agent behaviors.
• Preparing for scaled Korean content harvesting