AI agents and crypto payments: a match made in tech heaven
AI might not be plotting behind the scenes and screens to take over the world (hopefully), but it appears to be ready to participate in the economy, which is just as wild if you ask us. Over the past few years, we’ve become quite used to interacting with AI-powered chatbots and digital assistants, or using systems and tools that employ machine learning and AI technology in one way or another. In fact, some people have started to form bonds with AI programs like ChatGPT, training them to become their loyal virtual friends. That’s how far this innovative technology has come and how deeply it has integrated into our lives.
But having AI robots handle financial transactions autonomously, without any human intervention, is on a whole different level of futuristic development – and crypto seems to play a key role in this emerging trend. As far as crypto predictions go, no one really expected to see AI become a key player in the crypto economy, as most projections are centered around the potential price movements of different digital assets, such as Dogecoin prediction data and other noteworthy coins. But there’s a lot more going on in the crypto market than prices rising and falling under the influence of internal and external factors, and the intersection of crypto payments with AI technology proves that abundantly.
AI agents as the dominant force in crypto payments
So far, concerns regarding crypto adoption centered around whether people would be willing to use this digital form of money as an alternative or alongside fiat. Now the narrative is shifting as AI agents have also entered the picture, and they might be the ones to dominate crypto payments in the future.
In early March, Changpeng Zhao, the founder and former CEO of Binance, took to X to express his thoughts on the matter. In his post, he stated that AI agents will make one million times more payments than people, and they will turn to crypto as their instrument of choice.
Several studies reached similar conclusions. For instance, a report by McKinsey found that AI agents could enable somewhere between $3 trillion to $5 trillion worth of commerce transactions for consumers all across the world by 2030.
The Bitcoin Policy Institute conducted an experiment where they asked 36 AI models to act as economic agents and choose the assets they prefer for transactions. All of them chose crypto, although they weren’t influenced in any way. The experiment included 9,072 different scenarios, and AI agents chose Bitcoin in 48.3% of them. So, Zhao’s prediction that AI agents will take the reins in crypto payments is not as far-fetched as some might think.
Why crypto?
The technology behind AI and crypto might be complicated, but the reason why AI agents are more likely to choose crypto over other types of currency is rather simple. Banks and financial institutions control the way money and financial instruments are issued and used through very strict policies, highly regulated systems, including stringent Know Your Customer (KYC) requirements. AI agents are automated software programs and, therefore, don’t meet the requirements imposed by banks to use traditional financial infrastructures and engage in economic activities.
Crypto, on the other hand, is based on a decentralized model underpinned by blockchain technology. There are no central entities that manage the flow of assets, but a network of nodes that ensure the validity and security of transactions. The lack of central controls, intermediaries, and regulatory restrictions makes crypto assets a lot more accessible than fiat money, so anyone, including AI agents, can use them by simply setting up a crypto wallet that can be accessed and operated with a private key.
Therefore, an AI model with a valid crypto wallet can transact digital currencies just like a human would. These transactions can take place without a human identity attached to them, ensuring a completely automated process.
Brian Armstrong, the CEO of cryptocurrency platform Coinbase, also touched upon the topic that Zhao addressed, saying that AI agents are restricted from using conventional financial systems due to KYC protocols, but there are no such issues with crypto. Armstrong also mentioned Coinbase’s recently launched Agentic Wallets, specifically designed to facilitate the agent-powered payments.
How can AI models have monetary preferences?
Although AI has been created to mimic human behavior, it doesn’t function independently, so it can’t be said that AI bots have a mind of their own. So, how can they develop preferences for certain financial instruments like crypto? Well, it all comes down to how these models were trained.
In the experiment run by the Bitcoin Policy Institute, the AI agents were fed data regarding the advantages and disadvantages of Bitcoin and digital currencies, as evidenced by the available corpus of studies on the subject. The agents were then presented with different financial situations, like long-term investing or making payments, and they were given the opportunity to choose how they want to approach each task.
The conclusion they’ve reached, based on the knowledge they had acquired, is that Bitcoin is the best option for saving and investing (chosen in 79% of the situations), while dollar-pegged stablecoins were the go-to option for payments (53%). So, without receiving any suggestions or advice, AI models created their own monetary strategy, with Bitcoin as a savings tool and stablecoins as a spending solution, similar to how gold and silver have been used in the past.
Some argue that the BPI study is not accurate because AI models could not reach an objective conclusion since they were trained on human-written content, which is largely created by crypto supporters. That would explain the preference for digital assets over fiat currencies. Furthermore, this was a controlled experiment, so the results may be completely different in a real-world environment.
However, with various crypto companies already involved in the construction of agent payments infrastructure, it’s pretty obvious that soon enough we’ll see AI models become active participants in commerce and financial transactions, and using crypto as the base asset for it.
