Anthropic’s Claude surpassed OpenAI’s ChatGPT in South Korea’s paid generative AI market for the first time in March 2026. According to data from Hankyung Aicel, a research unit of the Korea Economic Daily, Claude’s domestic credit card payments reached 49.5 billion won (approximately $35.7 million) last month, overtaking ChatGPT at 47.2 billion won. Google’s Gemini trailed at 13.7 billion won.

Payment-based market share: Claude 42.4%, ChatGPT 40.4%.

Three Months From One-Third to First Place

The speed of the shift is the story. In January 2026, Claude’s payment volume was 15.7 billion won, roughly one-third of ChatGPT’s 49.1 billion won. By March, Claude had tripled while ChatGPT stayed flat. KED Global reported that Korean-language support improvements and the introduction of won-denominated pricing plans in late 2025 laid the groundwork.

Enterprise Spending Drove the Flip

The growth was not consumer-led. According to the Korea Economic Daily, the average corporate card payment for Claude in Q1 2026 was 181,000 won ($130), nearly three times the personal card average of 65,000 won ($47). Hankyung Aicel research head Hong Se-hwa attributed the gap to enterprise teams subscribing to Claude’s highest-tier plan (Max 20x, $200/month) for work use.

This enterprise concentration means Claude’s revenue per user in Korea is substantially higher than ChatGPT’s. ChatGPT maintains a larger user base in absolute terms, but Claude is capturing more spending per subscriber.

Why Korea Matters as a Signal

South Korea is the fifth-largest market for Claude usage globally, according to Anthropic’s Economic Index. The country’s enterprise sector is dominated by large conglomerates (Samsung, SK, LG, Hyundai) with significant AI procurement budgets and strict compliance requirements. Claude’s positioning as an enterprise-first, safety-focused model aligns with the governance expectations of Korean corporate buyers.

The Korea result is also notable because it happened without Claude leading in consumer awareness. ChatGPT remains the default name recognition brand in Korea’s consumer AI market. Claude’s payment overtake came entirely from the enterprise segment, suggesting that B2B adoption dynamics can outrun consumer brand dominance in paid AI markets.

What This Tells Us About the Enterprise AI Race

The Korea data point supports a thesis that has been building across multiple markets: enterprise buyers are not defaulting to the consumer-leading AI brand. They are evaluating on capability, compliance, and pricing fit. Claude’s growth in Korea, Japan (via the NEC partnership announced April 23), and globally through products like Claude Code and Managed Agents suggests Anthropic’s enterprise-first strategy is converting to revenue faster than its consumer metrics would suggest.

Google’s Gemini, at 13.7 billion won in Korea, remains a distant third in paid spending despite Google’s infrastructure advantages and the Gemini Enterprise Agent Platform rebrand at Cloud Next 2026 last week.