MCP for Investing: 8 Ways to Use AI with a Freedom24 Account

Freedom24 has opened its proprietary Tradernet brokerage infrastructure to AI applications through the Model Context Protocol (MCP), allowing clients to connect their investment accounts to AI tools including Claude Code, Cursor and Codex.

The connection gives an AI assistant access to selected portfolio, market-data, trading, reporting and alert functions. What does that mean in practice? Investors can use AI assistants to analyse portfolios, retrieve market data, review trading history, prepare order instructions and coordinate information across connected applications.

The shift is not simply from clicking through a trading platform to asking a chatbot the same questions. It is a move from navigating financial applications to expressing an intention. An investor can describe what they want to understand, monitor or prepare, and an AI assistant can bring together the relevant data and tools. In this context, agentic AI refers to AI software that can call external tools and coordinate multi-step tasks for a user; it does not necessarily mean autonomous trading.

That idea fits into a broader shift in brokerage technology, as more trading platforms and brokers begin connecting account data and trading workflows to AI through MCP. Freedom24 has described AI agents as a new infrastructure layer for its financial ecosystem, with the longer-term goal of allowing clients to express what they want and having AI coordinate the services needed to make it happen.

MCP for investing builds on that idea by providing a standard way for AI applications to connect to external tools and data, including selected brokerage account functions. An MCP-compatible AI client can therefore connect to Tradernet alongside separate MCP-enabled services such as a notes app, spreadsheet or messaging service.

Some of the examples below use additional services and require users to configure those connections themselves. They are not all ready-made Freedom24 features. Instead, they show what becomes possible when an investment account becomes part of a wider AI environment.

“The shift we are seeing is from navigating financial applications to expressing intent. Instead of working through separate screens and tools, an investor can tell an AI assistant what they want to understand or accomplish, and the technology can connect the right data and services behind the scenes. Opening Tradernet to MCP is an important step in that direction. We see the investment account becoming part of a wider personal AI environment, while the investor remains in control of the decisions that matter”, notes Vitaliy Zvonar, Chief Product Officer at Freedom24.

What can MCP do with an investment account?

1. Understand a portfolio without navigating the account

Instead of opening several screens and calculating exposures manually, an investor can simply ask an AI assistant a question such as:

“How much of my portfolio is invested in US technology companies, and which currencies am I exposed to?”

The assistant can retrieve portfolio information and turn the underlying account data into a concise answer.

The same approach can be used for more routine questions: how many positions are currently open, which holdings account for most of the portfolio, or how a portfolio is divided across sectors, markets or currencies.

The benefit is less about replacing the portfolio screen than making the information available through a more flexible interface.

2. Turn market data into questions

Tradernet's MCP toolset includes access to quotes and historical market data, allowing an AI assistant to retrieve information about securities and combine it with the investor's account context.

An investor could ask:

“How has this stock moved over the past six months?”

Or:

“Compare the recent performance of the three technology stocks in my portfolio.”

That makes market data conversational rather than forcing investors to move between multiple screens.

The AI is still working with underlying market data; it is simply providing another interface through which the investor can query and organize it. Its response should not be treated as independent investment advice.

3. Prepare a portfolio rebalancing

An investor who aims to keep 60% of a portfolio in ETFs and 40% in individual stocks can ask the assistant how far the current allocation has moved from that target.

The assistant can calculate the difference and draft trade instructions that would move the portfolio closer to the investor's stated allocation, for review.

That does not mean the investor has handed over the decision. Freedom24 says each trading order requires client confirmation. For Codex, Tradernet's optional plugin also adds approval prompts for write, trading and authentication tools; a basic MCP-only setup may not include those client-side prompts by default.

The important distinction is between preparing an action and approving it. MCP can make the first step conversational, while the investor remains responsible for reviewing what is ultimately submitted.

4. Turn trading history into a useful report

A brokerage account can contain months or years of orders, trades and reports that are difficult to review at a glance. Tradernet's MCP tools include access to current orders, order history, trade history and broker reports.

An investor could ask:

“Summarize my trades over the past year and group them by security.”

Or:

“Show me the positions I opened and closed most often.”

The assistant can organize the existing account information into a readable summary or table that can be used for personal records or shared with an accountant.

The value here is not a new source of financial information. It is the ability to transform existing account data into a format that is easier to review.

5. Make alerts part of a wider workflow

Tradernet's MCP tools also include price and other alerts.

An investor can use natural language to describe what they want to monitor:

“Alert me if this stock falls below the level I'm watching.”

The potential extends beyond a simple alert.

When the brokerage account is connected to other MCP-enabled services, an alert could become part of a wider workflow. The investor might record the event in a spreadsheet, add a note to an investment journal or pass the information to another application.

The exact workflow depends on the services connected to the AI assistant and the permissions granted to them. The principle is that an alert no longer has to end inside the brokerage platform.

6. Create a morning investment brief

An AI assistant connected to the investment account can also be used to prepare a short briefing at the start of the day.

The investor could ask for a summary of overnight moves in the securities they hold, the latest prices and any account alerts that need attention. If the same AI application is connected to an email or messaging service, the output can be passed there as well.

This changes the role of the brokerage account. Instead of being a place the investor has to visit each morning, it becomes a source of information that another application can bring into an existing routine.

A scheduled workflow would still require the relevant AI client and connected services to support scheduling or automation; it should not be interpreted as a standard Freedom24 feature.

7. Keep an investment journal with Notion or Google Sheets

One of the most practical combinations is an investment account with a notes app or spreadsheet.

An investor can connect Tradernet to a service such as Notion and ask the assistant to record a new trade alongside the reason for making it. Notion's MCP integration currently allows compatible AI applications to read from and write to a workspace, subject to the permissions of the connected account.

The same idea can be extended to Google Sheets. Google currently offers a remote MCP server for Sheets in developer preview, allowing compatible AI agents to read spreadsheet data and make changes such as updating values and formulas.

Over time, that can create a searchable record of how and why a portfolio changed:

“Bought 20 shares because I wanted to increase exposure to the sector after reviewing the company's results.”

Months later, an investor could ask the assistant to compare the trading history with those notes and identify patterns in their own decision-making.

The result is more than a record of transactions. It can become a record of the investor's reasoning.

8. Build a personal dashboard, or connect the portfolio to the home

The most experimental use case is also the one that best demonstrates the wider potential of MCP.

With an AI coding assistant such as Codex or Cursor, an investor can describe the dashboard they want, for instance, a page showing holdings by sector, country, currency and portfolio weight, and have the assistant generate much of the code needed to build it.

The same principle can extend beyond a traditional screen.

Home Assistant has an official MCP server that allows compatible AI applications to access selected home information and control exposed devices. Its documentation includes examples of using AI clients to control lights and other connected devices.

That means a technically minded investor could build a setup in which a connected lamp reacts when a monitored stock reaches a chosen alert level, or create a voice interface that answers a question such as:

“How is my portfolio today?”

These are self-built projects, not standard Freedom24 features, and they require additional configuration. But they demonstrate the larger idea behind MCP: the brokerage account can become one service among many that an AI assistant coordinates.

How to Use MCP Safely with an Investment Account

Connecting an investment account to an AI application creates a different set of responsibilities from simply using a chatbot to discuss markets.

The first is access. Tradernet's current documentation says API keys and secrets should be stored securely and not shared or published in repositories.

The second is permissions. An AI assistant connected to several services can potentially access far more information than the brokerage account alone. Notion, for example, notes that MCP tools operate with the permissions of the connected account. Investors should also consider how connected AI applications and third-party services handle account information, including data access, storage and retention.

The third is human review. AI assistants can misunderstand a request, make a calculation error or interpret an instruction incorrectly. Any prepared order should therefore be checked carefully for the security, direction, quantity and price before approval.

It is also worth considering where information is displayed. A voice assistant in a shared room, for example, may reveal account information to other people nearby.

And while MCP can make investment workflows more flexible, it does not turn an AI assistant into an investment adviser. The investor remains responsible for financial decisions.

MCP changes the role of the investment account from a platform investors visit to a service that AI assistants can interact with. Through Tradernet, investors can analyse portfolios, prepare allocation adjustments, review trading history and connect account information to other applications.

Freedom24's integration demonstrates how agentic AI can bring financial data and services together around an investor's objectives. The technology offers a more conversational and connected way to interact with brokerage infrastructure, while investment decisions remain with the investor.