The brokerage's Model Context Protocol server now connects directly to ChatGPT, enabling portfolio analysis and trade execution across stocks, options, crypto, and bonds through natural language conversations.
Public.com just made it possible to manage your brokerage account by talking to ChatGPT. The investing platform has integrated its Model Context Protocol (MCP) server with OpenAI’s chatbot, creating a direct pipeline between conversational AI and actual trade execution.
In English: you can now ask ChatGPT to analyze your portfolio, research stocks, and even place trades on Public, all through a chat interface. Every trade still requires your explicit confirmation before it goes through.
How the integration actually works
The connection runs on MCP, a framework OpenAI has been rolling out to let ChatGPT communicate with third-party services. To use it, Public users need to enable Developer Mode in ChatGPT and add the Public MCP connector. Once that’s set up, the AI can pull real-time data from your account and interact with it.
The integration supports trading across multiple asset classes: stocks, ETFs, options, bonds, and cryptocurrencies. Crypto services on the platform are facilitated through Zero Hash LLC.
Public has made it clear that users maintain full accountability for any actions taken on their behalf through AI agents. The architecture is designed so that while AI can suggest, research, and prepare trades, the human always has final say.
Public’s AI trajectory and the broader MCP trend
This isn’t Public’s first foray into AI-powered investing tools. The company launched “Alpha,” a GPT-4-based research tool, back in May 2023. That tool used OpenAI plugins to enhance investment research capabilities, giving users AI-driven analysis of markets and individual securities.
Public isn’t operating in a vacuum here. The MCP framework is becoming something of an industry standard for connecting AI to financial services. Morningstar, PitchBook, Alpaca, and Robinhood have all been integrating similar MCP functionality.
What this means for investors
There’s a real risk worth flagging. AI-assisted trading could create a false sense of confidence. ChatGPT is impressive at synthesizing information, but it’s not a financial advisor, and its responses can be wrong. The confirmation step helps prevent accidental trades, but it doesn’t prevent bad decisions that a user enthusiastically approves because the AI presented them convincingly.
