OpenAI launched a preliminary version of a
personal finance function in
ChatGPT that
allows Pro users in the United States to connect their bank accounts and consult their financial activity from the chatbot. The novelty incorporates a panel with recent transactions and the option to make personalized inquiries about the user’s financial context.
The tool allows you to connect accounts from more than
12,000 financial institutions to
ChatGPT. After linking, the system analyzes the data and answers queries ranging from reviewing recent expenses to developing
savings or
investment plans, with information on mortgages, debts, purchasing goals, or savings strategies. According to
OpenAI, the function helps detect consumption patterns and assist in relevant decisions with personalized responses.
The function also includes a collaboration with
Intuit, a financial software company. Based on that agreement, users can schedule sessions with local tax experts through
ChatGPT, with access to professional advice on taxes and personal finances.
OpenAI assured that ChatGPT does not see full account numbers nor can it modify banking information.
Accessing banking information from a chatbot reignited the debate on privacy and data security. OpenAI assured that the connection with bank accounts is secure, that the chatbot does not access full account numbers nor can it make modifications. Access is limited to balances, transactions, investments and liabilities, necessary to answer financial inquiries.
However, there is a possibility that
OpenAI uses this data to train its models if the user keeps the “
Improve the model for everyone” option enabled, which is usually enabled by default. Although the company indicated that it is possible to disable this function, it did not detail other potential uses of the information collected.
We recommend reading:What happens to your data in ChatGPT? Privacy risks and how to protect them
Critics and security experts, such as
Gang Wang, a computer science professor at the
University of Illinois, warned CNN that sharing financial information with AI systems may involve additional risks. A leak, security breach, or cyberattack could expose sensitive data and facilitate
phishing campaigns or fraud based on user transactions.
Security experts warned that a leak or attack could expose sensitive data and facilitate fraud or phishing.
ChatGPT and User Trust
Despite the doubts,
OpenAI reported that more than
200 million people use
ChatGPT each month to ask for help on topics such as budgeting, investment strategies, and financial planning. The company is betting that integration with banks and tax advisors will increase the platform’s utility in a competitive market.
However, OpenAI’s track record in terms of transparency and data handling was subject to scrutiny. The company repeatedly modified its policies and statements regarding data usage, while facing pressure for profitability and rumors about a possible IPO. In addition, CEO Sam Altman was involved in public legal disputes, which fueled questions about the reliability of the company’s promises.
Risks and Recommendations for Users
The access of
ChatGPT to bank accounts represents an advance in the integration of
artificial intelligence and personal finances, but opens new questions about data protection and risk management. Those who evaluate trying the function should review the data usage settings, consider the privacy implications and take extreme care against possible fraud or leaks. The option “Improve the model for everyone” could enable the use of data to train models if the user does not disable it.
OpenAI plans to expand the tool in the future, with openness to more users and new features. For now, the function is only available to
ChatGPT Pro subscribers in
United States.
