OpenAI is moving ChatGPT closer to the center of everyday money management by letting users connect real financial accounts. The question is whether people will trust an AI assistant with the most sensitive data in their lives.
OpenAI has started rolling out a personal finance experience inside ChatGPT, and this is not just another chatbot feature. It is a move into the place where budgeting apps, bank dashboards, robo-advisers and spreadsheet habits have been fighting for years: the daily view of your money.
According to OpenAI's May 15 announcement, the preview is available first to ChatGPT Pro users in the U.S. on web and iOS. Users can connect bank accounts, credit cards, investment accounts and liabilities through Plaid, with Intuit support planned. Once the accounts are linked, ChatGPT can show spending, subscriptions, upcoming payments, portfolio performance and broader financial patterns in one place.
That sounds simple enough. But the strategic shift is much bigger. ChatGPT is not only answering finance questions from memory or general knowledge anymore. It can now ground answers in actual balances, transactions, investments, debts and goals the user has shared. That turns it from a financial explainer into a financial interface.
OpenAI says more than 200 million people already come to ChatGPT each month with finance questions, including budgeting, investing, planning and comparing financial choices. Until now, the assistant could help with a savings plan or explain a mortgage tradeoff, but the user still had to provide the numbers. Most people do not have those numbers ready. Their financial life is scattered across checking accounts, cards, brokerage accounts, student loans, subscriptions and bills.
The new feature tries to remove that friction. A user can ask why spending rose this month, whether subscriptions are worth cutting, or how a planned purchase affects a savings goal. ChatGPT can then look at the connected data and respond with context. That is exactly where traditional budgeting apps have often struggled. They can categorize transactions and show charts, but they rarely explain what to do next in language that feels tailored.
This is why Plaid matters. OpenAI says the feature supports more than 12,000 financial institutions through Plaid. That gives ChatGPT the same kind of account-linking infrastructure that powers many fintech apps consumers already use, but with the conversational layer sitting on top. If this works well, the app people open for writing, research and work may also become the place they check whether they can afford a vacation or accelerate a debt payment.
For fintech companies, that is not a small threat. Budgeting tools like Rocket Money, Monarch Money and YNAB have built businesses around helping users understand where money goes. Robo-advisers and brokerage apps have built interfaces around portfolio performance and planning. Banks have tried to keep customers inside their own apps. OpenAI is now testing whether the interface people actually want is not a finance app at all, but a general assistant that already knows the rest of their life.
Privacy is the main product risk
OpenAI is clearly aware that asking for financial access is different from asking for calendar access or shopping preferences. The company says ChatGPT cannot see full account numbers and cannot move money or make changes to accounts. It can access balances, transactions, investments and liabilities for analysis, but the control remains read-only.
Users can disconnect accounts, and OpenAI says synced account data will be deleted from its systems within 30 days after disconnection. The company also says finance conversations follow the same model training settings users choose across ChatGPT, financial memories can be viewed or deleted, and temporary chats will not access connected financial accounts.
Those controls matter, but they will not settle the trust question for everyone. Financial data is intimate in a way that most digital data is not. It shows income, debt, medical payments, rent stress, family obligations, political donations, gambling habits, job changes and private priorities. A spending ledger can become a biography without trying very hard.
There is also the question of where helpful budgeting ends and financial advice begins. OpenAI says ChatGPT can help users stay informed, but it is not a replacement for professional financial advice. That line will become harder to hold as answers get more personalized. A generic explanation of diversification is one thing. A response grounded in a user's portfolio, income, liabilities and goals can feel much closer to advice, even if the product says otherwise.
OpenAI is trying to manage that with stronger reasoning models. Finance conversations with connected accounts default to GPT-5.5 Thinking, and OpenAI says it worked with more than 50 finance professionals to evaluate the experience. Its internal benchmark gave GPT-5.5 Thinking a score of 79 out of 100 on challenging personal finance tasks, while GPT-5.5 Pro scored 82.5.
Benchmarks are useful, but money decisions are not just reasoning problems. They are emotional, regulated and sometimes irreversible. If ChatGPT misclassifies cash flow, misses a bill, misunderstands a tax issue or gives overconfident investment guidance, the cost may not be a bad answer. It may be a missed payment, a bad trade or a false sense of security.
The bigger market implication is clear. OpenAI wants ChatGPT to become a place where users do things, not just ask things. Personal finance is a serious test of that ambition because the value is obvious and the trust bar is high. If users connect their accounts and keep them connected, ChatGPT becomes much harder to replace. If they hesitate, it will show that even the most capable assistant still has to earn the right to sit between people and their money.
Also read: Korea's AI chip boom is starting to move bond markets. • AI exploit benchmarks put Mythos at the center of cyber startup strategy • Stanford shows enterprise AI gains depend on workflow design
