OpenAI connects ChatGPT to bank accounts via Plaid for financial advice

OpenAI showed a preview of a feature that allows connecting ChatGPT to bank accounts through Plaid. If a user agrees, the bot will be able to see their complete financial picture — from account balances to credit card debt — and provide more personalized advice.

How It Works

The connection is built through Plaid — one of the most common bridges between banking applications and financial services. According to OpenAI, about 12 thousand financial organizations work through this platform, including Schwab, Fidelity, Chase, Capital One and other major brands.

For ChatGPT, this is a step from general answers to budget questions to a mode where recommendations are based not on an abstract case, but on the user's actual accounts, spending, and debt burden.

The company explains the launch simply: more than 200 million people already come to ChatGPT every month with personal finance questions — from budgeting to finding ways to cut expenses.

Before, the bot could only reason about the data that a person manually inserted into the chat. Now OpenAI wants to remove this barrier and make the system see the current picture itself and be able to answer with context, not by guessing.

It's also important that OpenAI is not building the banking integration from scratch. Choosing Plaid lowers the implementation barrier: for many users, this is already a familiar authorization mechanism used in fintech apps to analyze spending, investments, and credit profile. From a product perspective, this is a smart move: instead of negotiating with each bank separately, OpenAI gets a ready-made channel to a large part of the market and can quickly test whether people actually need an AI assistant with access to their accounts.

• Connecting bank and brokerage accounts through Plaid
• Viewing balances, money flows and expense structure
• Taking credit card debt into account
• More accurate advice on budgeting and saving

For the user, the value of such a connection is that ChatGPT stops being a calculator in words. It will be able to see how much money comes in and goes out, where debt accumulates, which spending categories are inflating, and how well the advice actually matches the real situation.

This is no longer "how to save better in general," but an attempt to answer the question "what should I do with my accounts right now."

Where the Main Risk Is

This is exactly where the news becomes truly sensitive. Financial data is one of the most intimate types of digital information: it reveals not only income, but also habits, debts, investments, overdue payments, and the level of financial stress.

OpenAI emphasizes that the connection will be "secure," but the very essence of the feature means that the user must trust the chatbot with access to data that previously was stored only in the banking interface or in specialized fintech applications.

"More than 200 million people already come to

ChatGPT every month with financial questions."

The problem is not only privacy. Even if the connection and data storage are organized correctly, the question of recommendation quality remains. A financial assistant that sees a real balance and credit load becomes much more useful — but the error of such an assistant costs more. A misinterpreted transaction, an overly confident spending advice, or missed context can easily turn AI from a convenient helper into a source of extra risk.

This is why the launch in preview format makes sense: OpenAI is testing not just another feature, but the boundary of user trust.

What This Means

ChatGPT is increasingly turning from a universal chat into an interface on top of real services and personal data. Connecting bank accounts through Plaid can make money advice truly useful, but at the same time raises the bar of trust in OpenAI: now the user is evaluating not only the quality of the response, but also the willingness to let AI into the most sensitive part of their digital life.