[DigitalToday reporter Yoonseo Lee] ChatGPT is increasingly likely to become the first gateway for new cryptocurrency users.

Cointelegraph, a blockchain outlet, reported on June 18 that MoonPay has started supporting crypto purchase procedures inside ChatGPT. Coinbase’s Base ecosystem is building tools that let AI chatbots handle wallets and blockchain applications.

Until now, crypto users have typically entered through a sequence of signing up for an exchange, completing identity verification, creating a wallet, making a first purchase and using decentralised applications (DApps). The problem was that the process was too complex for beginners. Wallet addresses were long and complicated, and seed phrases and network fee structures also acted as barriers to entry.

As a result, conversational AI is increasingly likely to become the first point of contact, rather than an exchange or a wallet app. Users can ask an AI assistant what bitcoin is, how to buy it and how to send money abroad, and get guidance immediately. If AI is also linked to the actual transaction stage, the conversation with a chatbot itself could become the onboarding process.

The MCP gateway being pursued by Coinbase’s Base is also part of the same trend. MCP is a structure that lets AI systems connect in a standard way to external apps and databases, wallets and services. Instead of moving across multiple apps, users give instructions in natural language and the AI handles intermediate steps.

In that case, the central axis of the user experience could also change. So far, users have moved directly between exchanges, wallets and blockchains to carry out each step. Going forward, blockchains and wallets could keep running, but users may interact through conversation rather than control.

From a market perspective, a shift in the role of exchanges is also a point to watch. Even if exchanges handle liquidity and execution, AI could take the actual touchpoint with users. If such a structure takes hold, core competitiveness could shift from back-end technology to who controls user conversation flows and access routes.

But as convenience expands, trust risks also grow. Previously, users dealt directly with an exchange, a wallet or a blockchain network. In a chatbot-based structure, much of that trust shifts to an AI assistant. The problem is that users may follow AI suggestions as they are, without fully understanding not only blockchain technology but also how large language models (LLMs) work.

Security issues also remain. The possibilities cited include prompt injection that disrupts AI judgement, malicious plugins that exploit trust connections, and AI-generated conversations that make scams sound more convincing. Given that incorrect chatbot answers and transactions differ in their impact, the importance of safeguards is expected to be a task going forward.

At the same time, combining AI and blockchains could expand beyond a human interface into machine-to-machine finance. Developers are creating AI agents that handle subscription management, portfolio adjustments, payments and use of decentralised finance (DeFi) protocols under limited human involvement. Fully autonomous financial agents are still at an early stage, but the basic tools to support them have already been built.