As AI assistants can now not only write code and translate, but also help analyze bills and plan home purchases, the boundaries of artificial intelligence are shifting from "general capabilities" to "vertical expertise." Recently, OpenAI launched a personal finance tool for ChatGPT Pro users in the U.S., allowing users to connect over 12,000 financial institutions through Plaid. They can directly query consumption analysis, investment portfolios, subscription management, and even future financial planning within conversations—marking OpenAI's official expansion of AI capabilities into high-value, sensitive financial decision-making scenarios.
Core Features: Manage Finances Like a Chat
The core experience of the new financial tool can be summarized as "Connect - Insight - Plan":
| Feature Module | Specific Capabilities | Typical Scenarios |
|---|---|---|
| Account Connection | Integrated with 12,000+ institutions via Plaid, including Schwab, Fidelity, Chase, Robinhood, Amex, etc. | Synchronize bank cards, brokerage accounts, and credit card data with one click |
| Financial Dashboard | Visualize asset performance, spending trends, subscription expenses, and pending bills | Quickly understand where your money is going and how much you have left |
| Smart Q&A | Supports natural language questions like "Have I been spending more recently?" or "Help me create a 5-year home buying plan" | No need for technical jargon—ask financial questions like talking to a friend |
| Privacy Controls | Accounts can be disconnected at any time, and synchronized data will be automatically deleted within 30 days; manual deletion of "financial memories" is supported | Users have full control over data retention |
Users can access the feature through the sidebar "Finances" section or by typing @Finances in a conversation to start connecting their accounts. The system will guide them through the Plaid authorization process.
Strategic Background: Acquisition of the Hiro Team + GPT-5.5 Capabilities
This release is not an isolated action but the culmination of OpenAI's financial strategy:
- Talent Reserves: In April 2026, OpenAI acquired the core team of the personal finance startup Hiro, which had received support from top-tier VCs like Ribbit and General Catalyst. OpenAI clearly stated that the Hiro team's expertise in financial product design and compliance experience played a key role in launching this tool;
- Technical Foundation: The new tool relies deeply on GPT-5.5's contextual reasoning capabilities—financial questions often involve multiple accounts, cross-cycle scenarios, and complex conditions (e.g., "If I sell part of my stocks, what impact will it have on my tax liability and mortgage eligibility?"), requiring the model to accurately understand implicit logic and constraints;
- Professional Validation: OpenAI collaborated with financial experts to build a dedicated evaluation benchmark, optimizing the model's accuracy and compliance in financial scenarios.
Privacy Design: The "Minimization Principle" for Sensitive Data
In this high-sensitivity field of finance, privacy protection has become the top priority in product design:
- Data Isolation: Financial data is stored separately from regular chat history to prevent cross-contamination;
- User Control: Supports disconnecting connections by institution, and data is completely removed from the cloud within 30 days after deletion;
- Transparent Audit: Users can view and manage all "financial memories" (e.g., "User plans to retire in 2030") on the "Finances" page.
This design addresses regulatory requirements (such as the U.S. CFPB data rules and the EU Open Banking Directive) and aims to alleviate user concerns about "AI holding financial privacy."
Industry Trends: General Large Models Accelerating "Verticalization"
OpenAI's financial tool is not an exception but reflects industry consensus:
- Real Demand: Over 200 million users ask ChatGPT financial questions each month, proving that general models have already taken on the role of "informal financial advisors";
- Competitor Follow-up: Anthropic launched a health consultation tool, Perplexity introduced a financial research product based on Computer Agent, and competition in vertical markets is intensifying;
- Capability Boundaries: When users start asking "Which stock should I buy," the "disclaimer-style" responses of general chatbots are no longer sufficient, and specialized, traceable, and reasoning-based vertical tools have become an inevitable direction of development.
Challenges and Outlook
Despite its promising future, the tool still faces multiple challenges:
- Compliance Risks: The U.S. SEC has strict definitions regarding "investment advice." If the model's output is deemed "stock recommendations," it may trigger licensing requirements;
- Liability Clarification: If users make loss-making decisions based on AI advice, who bears the responsibility?
- Data Dependency: The value of the tool heavily depends on the breadth of account connections. If banks limit API access due to security concerns, the user experience will significantly deteriorate.
