Agentic AI and Its Role in Finance

October 25, 2025
October 25, 2025 Gridley Wright

Introduction

In the previous blog of “Business Implementation of AI for Efficiency and Consumers,” we broadly discussed how artificial intelligence has been implemented across the workplace by countless companies. This blog looks to build upon the prior while angling its perspective to hone in on the recent wave of more independent AI agents and their interactions with each other and cryptocurrency.

AI Agents

AI agents have been developed to act more autonomously than much of the preexisting artificial intelligence. At their core, AI agents are applications that are intended to execute upon tasks without relying upon much input from humans. Unlike its technological counterpart of generative AI, agents don’t merely create an idea, they execute upon it. Furthermore, instead of only learning during the training stage, agents can learn while it is in the process of working on a task (Marr).

One of the key components that allows these agents to operate autonomously is their capability to interact with each other. As outlined by Google’s recently released Agent2Agent (A2A) protocol, a new standard has been set so agents can interact with each other. As agents may have been developed using different models, having the shared protocol bridges the gap and allows for secure communication and collaboration between bots (Google). This new collaboration sets the ability for an agent to delegate tasks to another if it is unable to perform an action itself. The same aforementioned protocol also describes the Agent Card, which identifies agents, describes what they can do, and details the manner with which they can be interacted (Desai). These cards serve the same utility as a business card any human would use, and, with the standardized communication method, serves as the foundation to begin any interaction.

Incorporation of Crypto and the Blockchain

The blockchain and its applications is a driving force behind the movements of agentic AI. First, since agents are executing work on behalf of a company or an individual, they are typically compensated. The best option for agents, especially because they may transact with other agents, is creating a wallet with which they can exchange or store cryptocurrencies and non-fungible tokens (NFTs). As these autonomous agents are not considered legal persons, they cannot access traditional accounts such as banks, so decentralized wallets that do not require verification create a perfect match. Once an agent holds a wallet, they are able to engage in activity, as they can now make transactions that are verified using smart contracts on the blockchain and fulfill the payment with cryptocurrency, often the stablecoin USDC, backed by the US Dollar (Circle). Even if an agent does not have an initial balance and requires funds, they can utilize blockchain applications such as generating and selling NFTs, creating a starting income source.

Agents in Finance

Agentic AI is expected to be heavily implemented in the financial services sector. Much of this is expected to be used for operational purposes within a company. For investing, agents are expected to make more informed forecasts for securities using real-time data, but can take this a step further by actually conducting trades. Multiple agents may also me used in projects to address efficiencies and risks for the company (Citi). Despite the implementation for interior operations, anyone who uses these companies should expect to interact with these agents more frequently. As agents are to be implemented at all fronts of a business, users and customers will interact with these agents on websites and apps. For example, someone shopping online can make an inquiry with an agent, and it can not only interact with the shopper, but can also execute actions (such as purchases) based on the information at its disposal (Ranosys). Even at an individual level, if one gains access to an agent (which should be more accessible in the future), they can use it for personal gains, such as setting income targets and allowing the agent to execute crypto trades on their behalf (Luks).

Lucrisma 

By signing up with Lucrisma, you will be able to access your very own AI Financial Assistant. As you link more financial accounts to your Lucrisma portfolio tracker, whether it be bank accounts or crypto wallets, the assistant begins to interpret the data and learn more about your financial situation.  You can interact with the assistant via chats, and with the assistant’s more holistic understanding of your finances, it can develop a plan or give insights curated to your financials. If you have a question about some of your recent transactions or an existing loan, the assistant can access your tokenized data to provide you with a specific answer. Taken a step further, our AI will assist you in completing transactions of transfers between accounts or payments of bills, executing upon the goal you provide it. Just like agentic AI, Lucrisma is here to simplify your money management.