Crypto Executives Split Over Currency for AI Agents
AI agents are evolving from digital assistants into autonomous economic actors capable of booking services, managing schedules and executing transactions. That shift creates a need for money machines can hold and transfer without human intervention. Crypto wallets are emerging as potential “machine bank accounts” because conventional banking relies on human identity checks, but the industry remains divided over whether agents will settle through one dominant currency or navigate thousands of tokens behind the scenes.
CoinDesk reported on Aug. 29, 2026, that Animoca Brands Chairman Yat Siu expects 50 billion to 100 billion autonomous agents to transact online in the coming years. OKX Europe CEO Erald Ghoos predicted a neutral “super currency” that may be a stablecoin or an as-yet-unknown crypto asset, avoiding dependence on any single country. Juniper Research forecasts global agentic commerce could generate $1.5 trillion by 2030, while Citi said trust, identity and authorization safeguards must improve before autonomous payments scale.
All Coverage
1 original reportsThe Backstory
The history behind this eventAI Agents Emerge as Crypto’s Next Billion Users, Paying in Stablecoins
AI agents — software that can act on a user’s behalf — need payment rails to buy data, computing power and API access without human approval for every transaction. Coinbase’s x402 protocol turns the long-dormant HTTP 402 “Payment Required” status into a machine-to-machine payment standard. Stablecoins are emerging as a natural settlement asset because they can move globally around the clock and support micropayments, making agentic commerce a potentially significant new source of crypto adoption.
As of Aug. 23, 2026, Coinbase said x402 had processed more than 165 million payments worth about $50 million, with an average transaction of roughly 30 cents. About 99% of payments were settled in Circle’s USDC, while the network counted more than 480,000 active AI agents. Lincoln Murr, Coinbase’s head of AI product, compared the market with the “Napster/LimeWire era”: the technology works and usage is expanding, but standards, safeguards and viable business models are still being worked out.
Subscribe to Mark Radar Weekly
Every Friday, the week's strongest signals in your inbox. Unsubscribe anytime.
If you search news on Google, you can set Mark Radar as a preferred source—our coverage will show up more often in your results. Set as preferred source on Google →