ChainCatcher, citing The Block, reported that Ripple has announced a new developer toolkit for building “agentic payment” applications on the XRP Ledger (XRPL). The toolkit is designed to help developers create applications around machine-initiated payments, settlement flows and asset transactions, with AI agents able to carry out automated financial activity.
AI agents move into payment use cases
Ripple said AI agents are no longer only a future concept. According to the company, they are already taking part in real payment activities, including paying for compute costs, settling invoices and completing transactions without human intervention. As the use of AI agents expands, the market is accelerating the construction of payment infrastructure for machines, including wallets and stablecoin payment rails, so that AI systems can pay for services and trade assets autonomously.
The announcement came in the same week that Robinhood introduced a related initiative allowing users to try stock trading carried out by AI agents, with plans to expand to crypto assets in the future. MetaMask also released a non-custodial wallet solution aimed at AI agents. These developments all focus on practical financial transaction scenarios for AI agents, spanning stock trading, crypto assets and on-chain wallet infrastructure.
x402, XRP and RLUSD included in settlement support
Ripple noted that traditional payment systems are mainly built for processes initiated and approved by humans. AI agents, by contrast, require infrastructure with different characteristics, including fast settlement, predictable outcomes and the ability to complete payments without human approval. Ripple emphasized that its new toolkit also supports payment capabilities based on the x402 protocol and can settle transactions using XRP and Ripple USD (RLUSD).
At the same time, the IC3 team, made up of researchers from multiple universities, said that combining AI with blockchain can enable automated transactions, but AI agents still depend heavily on humans and underlying infrastructure. The team said these agents do not have full independence. That view adds an important boundary to the discussion around AI agent payments: automated transactions can be integrated into blockchain-based payment flows, but their operation still relies on human configuration, system support and infrastructure.