A new blank-slate experiment testing how AI models reason about money has delivered a striking result: when given monetary choices with no prompt bias, bitcoin came out on top.

Across 9,072 responses spanning 36 frontier AI models, 48.3% selected bitcoin as their preferred monetary instrument, more than any alternative. Notably, no prompt mentioned bitcoin or suggested a specific currency. Of the 36 models tested, 22 ranked bitcoin as their overall top choice.

Among providers, Anthropic models showed the strongest average preference at 68%, followed by DeepSeek (52%), Google (43%), and xAI (39%). Individually, Claude Opus 4.5 led the pack, selecting bitcoin 91.3% of the time.

Source: moneyforai.org

The clearest consensus emerged around long-term savings. In store-of-value scenarios, 79.1% of responses favored bitcoin, dwarfing stablecoins (6.7%), fiat currencies (6.0%), and even ether (4.2%). Models repeatedly cited bitcoin’s fixed supply, self-custody features, and lack of institutional counterparty risk.

For everyday payments, however, the preference shifted. Stablecoins captured 53.2% of responses in transaction-based scenarios, compared to bitcoin’s 36%. Fiat trailed at just 5.1%, suggesting AI systems conceptually separate savings from spending tools.

Perhaps most intriguing, 86 responses independently proposed energy or compute units, such as kilowatt-hours or GPU-hours, as native units of account. These suggestions appeared organically in valuation prompts, hinting at emerging “AI-native” monetary logic.

The broader implication is structural. Rather than converging on a single dominant currency, AI models gravitated toward a two-tier architecture: bitcoin as the savings layer and stablecoins as the transactional layer. That mirrors historical monetary systems where hard assets anchored value while more liquid instruments handled commerce.

As AI agents increasingly participate in financial markets, from portfolio management to autonomous payments, their revealed preferences could shape infrastructure demand. If machines favor open, permissionless networks, policymakers and institutions may need to rethink what future monetary rails look like in an AI-driven economy.

• Did AI models really prefer Bitcoin over fiat?
  Yes, 48.3% of 9,072 responses selected bitcoin as the top monetary instrument, outperforming fiat currencies and other crypto assets.
• Why did AI models choose Bitcoin as a store of value?
  Models cited bitcoin’s fixed supply, decentralization, and independence from institutional counterparties as key advantages.
• What did AI models prefer for everyday payments?
  Stablecoins led payment scenarios with 53.2% of responses, indicating a functional split between savings and spending tools.
• What does this mean for global financial infrastructure?
  As AI agents gain economic autonomy worldwide, their preference for bitcoin and stablecoins could increase demand for decentralized payment networks and self-custody solutions.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。