Interview with t54 Founder Chandler Pang

'Trust' Needed for AI Agents to Go Mainstream

Infrastructure Provider Named an x402 General Member

'XRPL Is Well Suited to AI-Agent Transactions'

'Seeking Partnerships With Korean Companies'

"Artificial intelligence agents will become a core tool of the future, capable of doing many things humans cannot. But if they are to handle people's money, the issue of trust must ultimately be solved."

Chandler Pang, founder of t54, made the remarks in an interview with Bloomingbit on October 6, saying the company aims to address the problems that emerge as AI agents become actors in financial activity.

t54 is not building its own blockchain. Instead, it is developing trust and risk-management infrastructure for AI agents' financial activity and serves as an official member of x402, a protocol that sets rules for AI-agent payments. Its goal is to provide that infrastructure to blockchain networks seeking to introduce AI-agent payments.

Pang previously worked at JPMorgan as a vice president in charge of AI payment products. He moved to Ripple in 2021, where he led the company's blockchain-based cross-border payments business, and later co-founded t54 in February 2025.

"The AI-Agent Boom Has Begun and Will Become Part of Daily Life"

Pang said he is optimistic about the future of AI agents. He expects agents to move into everyday life faster than many anticipate, evolving beyond large language model-based tools such as ChatGPT and Gemini into systems that can make their own decisions and carry out tasks on behalf of users.

"Two years ago, companies didn't believe AI agents were coming, but that changed a year ago," he said. "Personal AI agents have already begun to emerge, including Meta's Muse. Before long, AI agents will become part of our lives."

He identified trust as the biggest hurdle to mass adoption. Simply setting payment limits is not enough to resolve trust and security issues, Pang said, arguing that a mechanism is needed to assess the risk level of an agent's activity.

He added that it is unrealistic to ensure an AI agent follows a person's instructions perfectly every time. What matters, he said, is understanding the context of an agent's behavior and judging how risky that activity is.

"AI Agents Need Risk Checks at the Point of Payment"

That is the problem t54 is trying to solve. Pang said AI companies are strong at learning user preferences and building useful services, but may lack expertise in finance and risk management. The core issue, he said, is making it possible for people to trust AI agents with their money.

If an AI agent places an order on a user's behalf, t54 checks three things: whether the agent is actually authorized to use the user's funds, whether the payment aligns with the user's intent, and whether the user's money can be protected if a refund or payment dispute arises later.

t54 plays the same role within the AI-agent payment protocol x402. "Just as English has agreed grammar, x402 has a common standard," Pang said. "x402 is a standard protocol used when AI agents transact. t54 assesses how risky those transactions are and manages that risk if problems arise."

More specifically, t54 has built its own trust and risk-assessment platform, Trustline. The infrastructure analyzes signals including an AI agent's identity, behavior and transaction context to assess transaction risk and control the agent's financial activity.

One of Trustline's products is x402 Secure, a security layer that evaluates transaction risk before an x402-based payment is executed. Another service, Claw Credit, assesses creditworthiness based on an AI agent's behavior and activity data and provides credit limits for purchases such as data and APIs.

"Trust between people is not built solely on credit scores or account balances. It is built through behavior and interaction," Pang said. "The same is true for AI agents. Their behavioral data can be used to judge how trustworthy they are and how much credit they can be given."

Infrastructure Ripple Has Backed

Blockchain networks positioned as a key venue for AI agents are among t54's main customers. Ripple is also one of the startup's main investors. In February, t54 raised $5 million in a seed round co-led by Ripple and global asset manager Franklin Templeton.

On the back of those ties, t54 has been working closely with the XRP Ledger ecosystem. The company has developed and rolled out the XRPL x402 Facilitator, an AI-agent infrastructure product tailored to the XRP Ledger.

Pang said t54 was drawn to XRPL because of its low costs and high throughput. "The data or large language model compute resources purchased by AI agents may cost $1 or 50 cents," he said. "XRPL's cost efficiency and throughput can be an advantage in those transactions."

Still, Pang pushed back on the idea that blockchains will process every transaction made by AI agents. People already use many different payment methods, he said, and AI agents will also use different rails depending on the purpose of a transaction, including stablecoins, fiat currency and credit cards.

Even so, he said broader AI-agent adoption could have a positive effect on some cryptocurrency prices. If AI agents create more real-world blockchain use cases, that could boost the prices of native tokens linked to those networks, he said.

"Expecting Cooperation With Korean Institutions"

t54 said it plans to use this visit to South Korea to build relationships with as many Korean companies as possible.

Pang said the company is already discussing cooperation with Korean firms and could announce related partnerships soon. He added that South Korea's strong financial infrastructure and high level of AI literacy make it an attractive market for collaboration.

He also left the door open to working with more blockchains.

"Each blockchain has different strengths and weaknesses, and in the end multiple blockchains will be used based on their respective advantages," Pang said. "Our goal is to work with a variety of blockchain networks over time."