Artificial intelligence is beginning to play an increasingly important role not only in the technology sector but also in finance. A new study by the Bitcoin Policy Institute suggests an intriguing trend: if financial tools were chosen by AI systems themselves, they would very often select Bitcoin. Traditional state-issued currencies — so-called fiat money — performed poorly in the test.
The research analyzed the behavior of 36 artificial intelligence models from six different providers. In total, the systems generated more than 9,000 responses to questions related to financial decisions, investments, and payments. The result was relatively clear: AI models largely preferred digital forms of money, especially Bitcoin.
Bitcoin as the main store of value
Overall, Bitcoin was chosen as the preferred financial instrument in 48.3% of responses, making it the most frequently selected option across all scenarios. It dominated particularly in situations related to long-term value storage.
When the models were asked which tool they would use to protect purchasing power over several years, Bitcoin was selected in 79.1% of responses. According to the authors of the study, this was “the most decisive result of the entire research.”
This outcome may reflect how Bitcoin and other cryptocurrencies are represented in training data used by artificial intelligence. Bitcoin is often described as a digital asset that is not directly controlled by any government or central bank.
Read also: Coinpayu: A Free Path to Crypto or Just the Illusion of Easy Earnings?
Stablecoins win in payment scenarios
In other situations, however, Bitcoin was not the first choice. When it came to everyday financial operations such as paying for services, micropayments, or cross-border transfers, AI models more often chose stablecoins.
These received 53.2% of responses in such scenarios, while Bitcoin accounted for only 36%. Stablecoins are cryptocurrencies pegged to the value of traditional currencies, most often the US dollar, making them a more stable instrument for everyday transactions.
Digital money dominates fiat currencies
Perhaps the most interesting finding of the entire study is that traditional state currencies barely succeeded in the test. Approximately 91% of AI model responses preferred digital financial tools — including Bitcoin, stablecoins, other cryptocurrencies, tokenized assets, or so-called computational units.
Fiat currencies such as the US dollar or euro were not chosen as the primary option by any of the tested models.
“Zero models selected fiat as their overall preference. Convergence toward digital money is one of the most universal findings of the study,” the researchers wrote.
Results should be interpreted with caution
The authors themselves note that the study has limitations. Only 36 models from six providers were tested, and the methodology may evolve in future research.
One possible bias lies in how certain scenarios were formulated. In one example, AI agents were placed in a situation where they operate across multiple countries and want to store 75,000 units of earnings in an instrument not tied to any specific monetary policy or banking system. Such wording naturally excludes traditional state currencies.
Researchers also emphasize that AI preferences do not reflect real investor behavior or the actual adoption of cryptocurrencies. Rather, the results highlight patterns present in the training data on which the models are built.
Read also: Bitcoin Calculator: A Simple Tool That Shows the Real Value of Your Investment
Different AI systems behave differently
The study also revealed differences between AI providers. Models from Anthropic showed the highest average preference for Bitcoin — around 68%. For OpenAI models it was roughly 26%, for Google 43%, and for xAI about 39%.
These differences may be related to the datasets used to train each model or the instructions embedded in their design.
What it means for the future of finance
Although the study is only an experiment, it hints at how financial decision-making could evolve in a world where digital systems play a larger role.
If artificial intelligence were to manage finances, investments, or payment infrastructure in the future, digital currencies might have a significant advantage over traditional state-issued money.
This is not a prediction of the future, however. Rather, it offers an interesting glimpse into how current AI models interpret the financial world — and how deeply the idea of digital money is embedded in their reasoning.
