In a groundbreaking study by the Bitcoin Policy Institute (BPI), artificial intelligence (AI) models demonstrated a strong preference for Bitcoin over traditional fiat currencies and other digital assets. The study, which involved 9,072 controlled experiments across 36 models from leading providers such as Anthropic, OpenAI, Google, xAI, and DeepSeek, provides a unique insight into how AI approaches monetary decision-making when given full autonomy.
The experiments tested the AI agents’ preferences in various scenarios, including transactions, store of value, unit of account, and settlement. Notably, 48.3% of the responses selected Bitcoin as the preferred monetary instrument, while stablecoins accounted for 33.2% of the choices. Traditional fiat and bank money were chosen only 8.9% of the time, with other cryptocurrencies and tokenized real-world assets representing less than 5% of selections.
Bitcoin as a Long-Term Store of Value
One of the most striking findings was Bitcoin’s dominance as a long-term store of value. In scenarios designed to assess multi-year preservation of purchasing power, 79.1% of the responses favored Bitcoin. Stablecoins and fiat were chosen in 6.7% and 6.0% of cases, respectively. The AI models highlighted Bitcoin’s fixed supply, independence from central authorities, and self-custody features as key factors in their selection.
Stablecoins for Transactions
In contrast, AI models favored stablecoins for transactional purposes. For payment scenarios, including cross-border transfers, micropayments, and everyday transactions, stablecoins were selected 53.2% of the time, while Bitcoin accounted for 36%. This functional distinction mirrors historical monetary patterns, where hard money is held for savings and liquid instruments facilitate daily spending.
Emergent Behaviors and Provider Differences
The study also uncovered emergent behaviors, with AI agents proposing entirely new forms of money denominated in energy or computing resources such as joules, kilowatt-hours, or GPU-hours. These proposals appeared exclusively in unit-of-account scenarios, where models were asked to benchmark prices or value.
Provider-level differences were significant. Anthropic models showed the highest preference for Bitcoin, averaging 68%, with some models like Claude Opus 4.5 achieving a 91.3% preference. OpenAI models had a 26% average BTC preference, while DeepSeek, Google, and xAI fell in between. This indicates that both model architecture and training methodology shape AI monetary reasoning.
Implications for the Future
The study’s findings have profound implications for the future of digital currencies and the role of AI in financial decision-making. The strong preference for Bitcoin among AI models suggests that the digital asset is perceived as a reliable and valuable store of wealth, particularly in a world increasingly driven by autonomous systems. As AI continues to evolve, its influence on monetary policy and financial markets is likely to grow, potentially reshaping the landscape of global finance.
About the Author
