How developers solve the financial limitations of AI with blockchain

One of the limitations of AI, and by extension of robots, is their inability to manage transactions autonomously. To tackle this, developers now integrate AI with blockchain infrastructure, so that AI agents can be performed safely, manage digital assets and perform smart contracts.

At Eth Denver 2025, Decrypt Speaked with experts from the industry of the Coinbase Developer Platform, AI and Robotics Developers OpenMind and Robonomics to investigate how AI-driven humanoid robots and other systems use blockchain to function as independent economic agents.

“Out of the box, AI has no interaction with the blockchain at all” Decrypt. “If you want it to have a wallet, send money or transfer money to different places, you can’t. It cannot register for a bank account either. This is an important limitation because in many cases AI needs the possibility to handle. “

The Coinbase developer -platform recognized this gap and developed a framework called Agent Kit, which the company launched in November, Dalal said. Agent Kit can get AI a wallet, so that he can perform any action on chains that a person can do.

Dalal called it one of the earliest use cases for AI agents and noted that AI can handle tasks, such as following the internet and performing things on behalf of users.

“So far, more than 2,000 developers have built at the top of Agent Kit, which contributes to more than $ 100 million in value,” he said. “One of the most popular use cases is following something on the chains and performing an action in response in being a ‘like this, do that’ automation. ‘

As the influence of AI grows in several industries, blockchain developers increasingly investigate the potential for crypto traders and investors. Yet Dalal Ai -Hallucinations recognized a problem in the industry.

“It can take actions or provide information that is not true,” he said. “When real money is involved, this problem becomes even more problematic.”

Dalal said that mechanisms that give users more control and supervision are necessary to ensure that AI can handle securely financial transactions without risking unintended actions.

“For example, if a transaction is under a certain dollar amount, the AI ​​can continue automatically. If it exceeds that threshold, the approval may be necessary by the user, “he said. “Likewise, if the AI ​​pays a new party for the first time, it may have to be on the white list, while later transactions do not require extra approval.”

Despite these worries, Jan Liphardt, founder of Open-source AI and Robotics Startup OpenMind, said AI and Robotics a unique and compelling use case for Crypto.

“Blockchains were built for machines, and they are a crucial part of the infrastructure needed for machines to work well with people,” Liphardt said Decrypt.

Liphardt, a professor of bio-engineering at Stanford University and emphasized the challenges with which AI-driven humanoid robots are confronted in people-oriented environments, such as banks and airlines, who are not prepared to accommodate them.

“We see incredible amounts of friction such as thinking machines trying to navigate through the human world, and a large part of it is a lack of identity,” he said. To reduce this friction, Open Mind designed IRIS to use multiple cryptocurrency portfolios to facilitate transactions.

“She has two wallets,” said Liphardt. “One is a straight Ethereum wallet, and the other is a coinbase wallet,” he said, adding that Iris keeps the private keys to the wallet and checks the balances every six seconds.

Balancing innovation with safety

As robots become part of human systems, questions appear about their identity, how they earn and manage money, how they deal with people, Liphardt said.

What rules arrange them and how those rules can be changed are also in the spirit, where Liphardt adds decentralized led by, are very suitable for taking on challenges, because they already support many of those use cases.

In addition to managing interaction with the real world, developers have turned to blockchain technology to enforce transparent and unchanging rules of administrative rules to tackle concerns about humanoid robots.

“That is why we write governance and rules of conduct on Ethereum,” Liphardt explained. “When a humanoid robot starts, it reads those rules. If someone asks:” Why should we trust this computer? ” You can easily point out the rules of governance – they are unchanging and publicly accessible.

Although this offers some guarantees, Liphardt acknowledges that it is not a complete solution.

To alleviate the fear of humanoid robots, influenced by films such as ‘The Terminator’, Liphardt noted that medical robots are smaller and approachable than those used in warehouses or dangerous institutions.

For Vitaly Bulatov, evangelist with robotics startup Robonomics, the developer of the Humanoid Robot Unitree G1, Blockchain technology offers means that smaller robotics developers may not have access.

“Although we can perform small LLMs on the robot itself, we probably need more calculation to make this cognitive layer completely possible,” Bulatov said to Decrypt. “In our case we are built on Polkadot, so we serve as an entrance gate to Polkadot itself. We use calculation providers on Polkadot to get extra computing power for this cognitive layer. “

Bulatov said that to integrate autonomous robots seamlessly into our economy, they must be independent agents who are able to manage transactions and contractual relationships.

“What we do is essentially an identity, a wallet and giving the ability to enter into contractual obligations with people,” he said.

The system records data – including error messages and videoofeeds – ACY Time A payment is being processed. According to Bulatov, this information is registered with a decentralized storage provider on Polkadot, which makes it possible to complete the task.

“In this way, no matter how large the autonomous system becomes – whether it is millions of robots or a very smart city – we can do so, on the basis of payment, the technical implementation takes place,” Bulatov said.

Published by Sebastian Sinclair