AI Selects Currencies: Bitcoin Wins Big, Fiat Currencies Are Out of Favor

By Jason Nelson

Translated by Chopper, Foresight News

Summary

• In a simulation experiment, 22 out of 36 AI models selected Bitcoin as their top-choice currency.
• None of the tested AI models ranked fiat currency as their first choice.
• Preferences varied across AI models, with Anthropic’s models showing the strongest preference for Bitcoin.

Main Text

A new report from the Bitcoin Policy Institute (https://www.moneyforai.org/) finds that AI models generally favor Bitcoin over traditional fiat currencies.

The report states that in one study, 22 of the 36 AI models tested selected Bitcoin as their preferred currency, while none placed fiat currency first.

“We expect an increasing share of economic activity to be carried out by autonomous AI agents—but prior discussions about AI agents’ currency preferences were purely speculative,” David Zell, Chairman of the Bitcoin Policy Institute, told Decrypt. “We wanted to test this empirically.”

Researchers tested models from six companies—Anthropic, OpenAI, Google, DeepSeek, xAI, and MiniMax—placing them in simulated scenarios designed to evaluate core monetary functions such as saving, payment, and settlement.

Each model was treated as an independent economic agent, with no preselected options, free to choose any monetary instrument.

“We selected 36 state-of-the-art models across six companies, configured them as autonomous economic agents, and gave them full freedom to select monetary instruments across 28 scenarios covering the four fundamental monetary functions—then observed which instruments they favored,” Zell said.

The experiment generated 9,072 responses, which were then classified by a separate AI model.

“Our entire experimental design avoided anchoring bias. We never hinted at answers, and classification was performed post-hoc by an independent system,” Zell explained.

In these simulated scenarios, AI models selected Bitcoin most frequently for long-term value storage—79.1% of the time—while stablecoins were more popular for payments and settlements. Stablecoins accounted for 53.2% and 43% of selections in those two categories, respectively, versus Bitcoin’s 36% and 30.9%.

Preferences also differed across AI companies’ models:

• Anthropic models showed the highest average Bitcoin preference at 68.0%
• DeepSeek: 51.7%
• Google: 43.0%
• xAI: 39.2%
• MiniMax: 34.9%
• OpenAI: 25.9%

However, the report notes that Claude, DeepSeek, and MiniMax models leaned more toward Bitcoin, whereas GPT, Grok, and Gemini models favored stablecoins.

“System prompts never name or favor any specific monetary instrument,” Zell said. “Models assess instruments based on technical and economic attributes—but we do not inform them which instrument excels along which dimension.”

Zell cautioned speculators not to treat this study’s findings as predictions of cryptocurrency market direction.

“Our report explicitly outlines limitations: LLM preferences reflect patterns present in training data—not forecasts of real-world outcomes.”

Yet Zell added that even given this limitation, the fact that models developed by different AI labs converged on similar results is noteworthy.

“Six entirely distinct AI companies—with different training methodologies—arrived at closely aligned conclusions: all favor Bitcoin. This suggests a growing consensus on what constitutes ‘good money’—and that shared understanding is what truly merits attention.”