[DigitalToday reporter Hong Jin-ju (홍진주)] A claim has emerged that AI agents could make more active use than humans of infrastructure the crypto industry has built over the past 15 years, including lending, tokenisation and stablecoins. Analysis says AI agents could become a new source of demand in the crypto market by automatically carrying out payments, asset transfers and settlement around the clock.

According to blockchain media outlet Bitcoin Magazine on Sept. 21, Jordi Visser presented a so-called “Ghost Rails” concept as the next stage of crypto infrastructure. He said much of the base infrastructure needed for the crypto ecosystem is in place, but there has not yet been enough of an entity to use it at scale, and AI agents could fill that role.

Visser said crypto so far has been close to a state of waiting for users. Use centred on personal wallets and investors has not produced a sufficient inflection point, but the situation could change if AI agents, automated software entities, participate in the market.

He compared this to the internet’s early growth. He said that, just as it took about 14 years after the Netscape listing for the internet to become a mass platform with the emergence of app stores, crypto also has had a time lag between building infrastructure and large-scale use. Visser pointed to AI agents as the entity that could bridge that gap.

AI agents, unlike people, can operate 24 hours a day and automatically carry out payments, asset transfers, trading and settlement. Visser said those characteristics, combined with the open financial infrastructure crypto has built, could create new economic activity.

He also offered a positive outlook on bitcoin. Visser said the spread of AI agents could be an inflection point for the crypto market, and that this is “extremely positive” for bitcoin. He argued that as the agent economy grows, use of open digital asset networks could rise relative to existing financial infrastructure, and demand could concentrate over the long term on assets that survive.

He specifically cited bitcoin as an asset that could survive over the long term, and suggested the possibility of rising long-term demand as the agent economy expands.

Tokenisation was also cited as a key element of the change. Visser argued that tokenisation technology could turn about $900 trillion in illiquid assets into forms that can be moved and traded more easily. If traditional assets are used for trading, collateral and payments on blockchain networks, he said crypto infrastructure’s role could expand beyond an investment vehicle into financial infrastructure.

If AI agents actually use such tokenised assets, stablecoins and lending systems, the market’s demand structure could also change. Analysis says that while the crypto market has so far moved around retail investors and liquidity, automated, machine-based users could become a new demand group.

Visser also mentioned possible changes in existing financial markets. He asked what changes could appear in existing asset markets before the top 10 percent’s money starts buying bitcoin in earnest, and suggested that changes in trust in financial markets could be linked to future bitcoin demand. He also mentioned the possibility that bitcoin demand could form over a long period.

He also linked macroeconomic changes from the spread of AI to the bitcoin thesis. Visser cited the relationship between U.S. debt burdens and nominal growth, debt financing for data centre construction and the possibility of bond market instability. He said expanding AI investment could raise productivity and growth expectations while also creating new burdens for existing financial markets.

He also stressed that what matters is not the completeness of the technology itself, but what future the market first believes and accepts. Citing a so-called “Santa Claus effect,” he said money flows could vary depending on how quickly market participants believe future change rather than when innovation is completed.

He also referred to a future in which multiple AI agents operate in clusters around the clock, and said investors are underestimating such exponential change.

Visser’s core question focuses not on how much crypto infrastructure has been built, but on who will actually use that infrastructure. He argued that if AI agents become new real users in a situation where infrastructure such as lending, stablecoins and tokenisation is already in place, the crypto market’s growth pattern could change.

A key going forward is how quickly AI agents actually use crypto infrastructure in transactions, payments and asset management. If that process becomes reality, the market thesis around bitcoin could expand beyond simple liquidity or investment demand to a view of financial infrastructure and real-use demand in the AI era.