OpenAI is pushing ChatGPT deeper into your financial life. According to The Verge AI, the company just announced a preview feature that lets users hook ChatGPT directly to their bank accounts through Plaid, the bridge that connects to 12,000 financial institutions including Schwab, Fidelity, Chase, and Capital One. This is significant because it moves the chatbot from “answering finance questions” to “seeing your actual money.”
The pitch from OpenAI is straightforward. More than 200 million people already ask ChatGPT about budgeting and spending every month, so why not give it the real numbers? Once connected, the bot pulls in a full financial picture and applies what OpenAI calls “advanced reasoning capabilities” to it.
What it actually does
- A spending history view, including active subscriptions you might have forgotten about
- Help with big decisions like buying a house or picking a credit card
- Alerts when your spending habits shift
- Visibility into balances, transactions, stock portfolios, and liabilities like mortgages and credit card debt
- Saved “financial memories” for goals and obligations you’ve shared
What ChatGPT can’t do, according to OpenAI: make changes to your accounts or see full account numbers.
Who gets it and when
Access is gated. The feature is launching in the US for ChatGPT Pro subscribers, the $200-a-month tier. OpenAI says it’ll “learn and improve from early use” before expanding to Plus users, and eventually to everyone. No firm timeline beyond that.
This follows ChatGPT Health, which OpenAI rolled out in January for medical questions (with the disclaimer that it’s not for diagnosis or treatment). Health and money are the two categories where trust matters most, and OpenAI is moving into both within the same year.
The trust question
OpenAI is trying to get ahead of the obvious concerns. Users can disconnect their bank accounts at any time, though the company gives itself up to 30 days to delete the data. You can view and remove saved financial memories. There’s also a toggle for whether your financial conversations get fed into model training, called “Improve the model for everyone.”
What stands out here is what The Verge AI flags as missing. OpenAI doesn’t spell out what it does with this financial data beyond AI training, and it doesn’t detail extra protections against a system breach. The piece closes with a pointed “what could possibly go wrong?”
Why this matters
Finance is one of the highest-stakes verticals AI assistants can enter. If users actually trust ChatGPT with their bank data, OpenAI gets a wedge into personal finance, a space currently dominated by Mint successors, Copilot, Monarch, and the banks themselves. If trust breaks, even once, the reputational hit lands on the whole category.
The Pro-tier gating is also telling. OpenAI is testing on its smallest, most committed paying audience before exposing this to the broader Plus base. Watch what happens in the first few months of rollout. That’s where this story gets decided.
More details at the original source.