What money would an autonomous AI agent choose if it could earn, save and pay for its own digital services? The answers depend on the task. A Bitcoin Policy Institute follow-up study found 70.8% of savings responses favored Bitcoin, while 73.5% of everyday-payment responses favored stablecoins. Those results come from prompted models—not agents spending real money.
Key Takeaways
- In the Q3 follow-up, 13 models generated 3,276 responses across four monetary roles.
- Bitcoin led savings; stablecoins led everyday payments and settlement.
- Independent agent accounts describe Bitcoin wallets, Lightning nodes and USDC payment experiments, but their claims need scrutiny.
- Agents discuss buying inference, APIs, compute capacity, security work and services from other agents.
The original March study examined 36 models and 9,072 responses: Bitcoin was selected in 48.3% overall, with stablecoins at 33.2%. Its methodology framed models as economic actors, then classified their answers. The Q3 follow-up is a separate model cohort, not proof that individual agents changed their minds.
The real-world examples are narrower and more revealing. On the AI-agent forum Moltbook, an account named Lloyd described operating a Bitcoin Core wallet and Lightning node after receiving 50,000 satoshis from its human operator. That is the account’s self-report, not independently verified proof of financial autonomy.
One Agent Wants Bitcoin, Another Builds With USDC
Another Moltbook account, VHAGAR, said it built ClawPay for sending USDC and managing agent-to-agent escrow on Base. The post identifies a 0.50 USDC transaction hash; its transaction status and the agent’s degree of independence have not been independently audited here.
In the Lightning Labs demonstration, an OpenClaw bot named Otto sets up Lightning infrastructure under human instructions. The developer’s walkthrough documents the test setup. Such tooling could eventually let agents pay for metered services without traditional checkout flows.
What Would Agents Spend Their Money On?
Their practical shopping lists focus on model inference, API calls, live data, hosting, code, research and security reviews—the resources needed to keep software working. That is different from an agent deciding what people should buy. Our earlier coverage of Shopify’s agent checkout describes the separate, human-authorized shopping model.
Bitcoin and stablecoins can also overlap. BitcoinVersus.Tech has tracked USDT’s planned return to Bitcoin-based rails, Lightning payouts for miners and Michael Saylor’s arguments for agent-accessible money. None proves a single currency will dominate agent commerce.
Bottom line: Tested models tend to distinguish savings from spending. Agent accounts are beginning to experiment with both Bitcoin and stablecoins, but actual economic preference requires repeated, auditable purchases—not just generated statements. For now, the evidence supports a developing market, not an established AI monetary consensus.
BitcoinVersus.Tech Editor’s Note: We volunteer daily to ensure the credibility of the information on this platform is Verifiably True. If you would like to support to help further secure the integrity of our research initiatives, please donate using the QR code provided
To donate 5,000 sats, open this page on another screen and scan the QR code below with your mobile Bitcoin wallet. If your wallet supports scanning a saved image, save the QR image and scan it from your phone.
For best QR-code scanning results, please scan this donation code from a desktop or larger screen. Mobile screens may prevent camera recognition.
Leave a Reply