'Too much data to read' — Bithumb builds AI eyes for crypto investors

"The biggest change in the digital asset market over the past few years has not been the price of bitcoin — it has been the structure of the market itself. With institutions entering and data exploding, what investors need now is not more information but something that can read the market for them."

Kim Yun-gi, head of Bithumb's Trading Research Institute, made the remarks Thursday at the 2026 Digital Asset AI Summit held at Sofitel Ambassador Seoul in Songpa-gu, Seoul, saying the market had grown so large that it had become harder, not easier, for investors to make sense of it.

The digital asset market has expanded rapidly as ETF launches, the spread of custody services and digital asset treasury operations, and the entry of traditional financial institutions such as securities firms have all broadened its reach. At the same time, a torrent of on-chain data, news and disclosures has made it increasingly difficult for retail investors to track and interpret every signal on their own.

To address the information overload its users face, Bithumb recently launched Bithumb AI, developed jointly with Surf AI.

Bithumb AI comprises four services: AI Asset Signal, which provides per-asset scores every hour; AI Asset Analysis, which aggregates on-chain data to produce in-depth reports every four hours; AI Recommended Questions, which refreshes answers to frequently asked questions — such as the causes of market declines — every 30 minutes; and AI Daily Market Update, which compiles key market issues and upcoming events.

It took about five months for Bithumb and Surf AI to develop an exchange-specific AI service. Kim said the team was often asked why a dedicated partner was needed when tools like ChatGPT or Claude already performed so well. "General-purpose AI models understand language, but they do not understand the digital asset market," he said. What was needed, he explained, was a specialist firm that could connect to real-time crypto data while also meeting an exchange's verification standards.

Surf AI is a market intelligence platform specializing in the digital asset market. It collects and analyzes prices, trading volumes, news and social media sentiment from more than 40 blockchains and over 200 data sources.

For retail investors, it offers AI research tools to explore relevant information; for businesses, it provides customized solutions and features that can integrate market analysis into their workflows. The platform has more than 600,000 users worldwide, including about 50,000 in South Korea.

Kim identified personalization as Bithumb AI's next goal — moving beyond delivering the same market information to all users and toward filtering and presenting information tailored to each person's portfolio and investment style.

Bithumb is also exploring a system that would analyze users' trading histories to identify individual investment characteristics, including habits, entry and exit timing, holding periods and profit-and-loss patterns. A service allowing users to converse with the AI to check market information and execute trades is also under consideration.

The focus, however, remains on helping investors organize vast amounts of information and work through the process of forming a judgment — not on having AI make investment decisions for them. "AI is not there to predict on your behalf," Kim said. "What AI changes is the way investors arrive at a judgment."

As personalization services expand, the scope of customer data use and the means of protecting it are expected to become a central challenge. Surf AI's market intelligence platform is designed to analyze market data without accessing individual customer information.

Park Seo-hyun, Surf AI's Korea head, said the first thing financial firms check when considering AI adoption is whether the AI accesses customer data. "Even throughout our collaboration with Bithumb, Surf has never accessed Bithumb's customer data," she said.

At the summit, participants projected that as the volume of assets traded on-chain and the data generated from those transactions continues to grow, so too will the role of AI in analyzing them.

Ryan Lee, co-founder and CEO of Surf AI, said that in the long run, crypto would become the foundational infrastructure underpinning all asset transactions. "Even if users are not aware of the technology, transactions will flow naturally on top of it," he said.

Lee added that as more assets and companies adopt crypto and move on-chain, the data accumulating on blockchains will grow, making financial analysis increasingly complex. "Demand for tools capable of analyzing that data is also rising," he said.