Mastercard just paid $1.8 billion for a stablecoin startup. Read that as an admission.
While most of the crypto world is debating which blockchain becomes the default infrastructure for the agent economy, the incumbents aren't debating. They're buying. Which protocol wins doesn't really matter – no one cares what's running underneath. They just want it to work.
Agents need a home: a messaging layer, a command interface, somewhere users live. Telegram has 1 billion monthly active users. Developers have been building on it for years, the wallet infrastructure is in place, and the payment rails are there.
TON was built for exactly this intersection. Being native to Telegram means we're already inside the stack where agent activity is happening. The plumbing just needs to catch up.
Every agent framework in the space has been building its own wallet logic from scratch – private keys scattered across config files, nothing that travels between them, every team reinventing the same wheel. Meanwhile the payment protocols were piling up: x402 for stablecoin payments, Stripe's Machine Payments Protocol for micropayments, Google's Agent Payments Protocol with sixty-plus partners. All of them assumed the agent already had a wallet. None of them said where it lived.
MoonPay just answered that question with the Open Wallet Standard – a single seed phrase that works across eight major chains, TON included, the key itself never touching the agent. We contributed alongside PayPal, the Ethereum Foundation, the Solana Foundation, and more than a dozen others.
Setting up an agent today is still like setting up email in 1995: doable, but only if you love reading manuals. That pile just became shorter, thanks to the Open Wallet Standard.
In theory, agents will route to the cheapest rail. But giving an agent a credit card takes minutes, while stablecoin onboarding is more steps than most consumers will go out of their way for. For now.
McKinsey puts the global agentic commerce opportunity at $3 to $5 trillion by 2030. Everyone wants to know who captures that: AI labs, cloud providers, card networks, some new entrant nobody's heard of yet. Honestly, I wouldn't be worrying too much about it. Networks embedded where a billion people live every day don't need to win that argument. They're already the answer.
Agents will replace roles that real people currently fill – moderation, research, executive support, customer response. That conversation is overdue.
The $3 to $5 trillion is coming. The question is whether a handful of platforms own it or nobody does.
