An AI agent that pays for research needs a spending limit. One that helps choose an investment needs to understand what the customer wants, how much risk they can take and where its own authority ends.
t54 Labs and RiskX are preparing to test that next step. The companies have signed a memorandum of understanding to develop a proof of concept for AI-powered, personalized structured investing, combining t54’s identity and risk infrastructure with RiskX’s financial-product expertise.
The proposed workflow would help customers describe their investment objectives and risk preferences, then use agents to identify suitable instruments and compare alternatives. The announcement sets out a planned test; it does not announce an investment service customers can use today.
From Buying Services to Choosing Investments
Much of the early work in agent finance has focused on getting software to pay for things: data, computing power and other services needed to complete a task. Investment decisions introduce another layer of complexity. A transaction may fit an agent’s spending limit while still falling outside the customer’s investment objectives.
RiskX specializes in structured products, with capabilities covering design, valuation, risk management, hedging and trade automation. These products can be configured around different maturity dates, return conditions and exposure to losses, even when they reference the same underlying assets.
That makes personalization more involved than selecting a ticker. Two products linked to the same market can produce different outcomes because their terms differ. An agent would need to understand those differences alongside the customer’s requirements.
“Personalized investment begins with understanding a customer’s objectives and risk parameters.”
— Sergio Chan, Co-founder of t54 Labs
Chan said institutions also need to know who an agent represents and what it is authorized to do. The planned collaboration would explore how those controls support more personalized investment workflows.
“In this PoC, we plan to validate not only the AI’s ability to recommend financial products, but also the process through which it understands a customer’s investment objectives and risk parameters and identifies an appropriate structure.”
— Janghyuk Youn, CEO of RiskX
Checking the Decision Before Money Moves
t54’s contribution centers on Trustline, its agent risk and underwriting platform. According to the company, it evaluates an agent’s identity, the authority granted by its user, transaction intent, policy restrictions, historical behavior and counterparty risk before a transaction proceeds.
For an investment workflow, the aim is to keep agents within defined permissions while retaining oversight over their actions. Being allowed to research a product would not automatically mean being allowed to commit the customer’s money.
That follows the argument t54 co-founder Chandler Fang made in his earlier Blockster opinion piece on trusting agents with money: financial platforms need to establish whether a particular transaction matches the authority and intent behind it.
The same distinction runs through XDC’s new identity integration with Concordium. Verifying the owner behind an agent is one part of the process. Deciding what that agent may do with someone’s money is another.
Post from @t54aiView the original post on X↗
13 Million Transactions—and a Different Kind of Test
Founded in San Francisco in 2025 and backed by investors including Ripple and Franklin Templeton, t54 has been building payment and risk infrastructure across the XRP Ledger and RLUSD ecosystems.
The company now reports more than 13 million agentic transactions on XRPL. That follows the 10 million milestone Blockster covered on October 1.
In the announcement, t54 also cites more than 169,000 agents applying for credit lines through Claw Credit, with more than $649,000 in credit extended. Those figures describe its existing agent-finance activity; they are not customer or investment-volume figures for the proposed RiskX project.
Its Tidal Wallet separates the authority to hold funds, initiate transactions and approve risk. Claw Credit addresses funding, giving agents controlled spending power without requiring users to pre-fund every transaction or grant unrestricted wallet access.
The RiskX collaboration would extend that work into investment selection. The payment activity establishes a base to build on, but assessing an investment against a customer’s needs presents a different test.
What the Pilot Needs to Show
The announcement does not specify a launch date, customer eligibility, supported products or the exact approval steps before an investment could proceed.
For now, the companies’ objective is to test whether an agent can understand a customer’s requirements, identify an appropriate structure and operate within clear controls. A useful result would show how a recommendation was reached, which permissions applied and where a person needed to approve the next step.
That is the practical challenge ahead: making personalized investing easier to navigate while keeping the customer’s objectives and authority attached to every decision.
Reporting by Lidia Yadlos
