Inrupt embeds a personal finance assistant in your own app, without handing customer data to AI vendors.
Source: Bloomberg, Sept 22, 2026; McKinsey & Company, Global Banking Annual Review 2025, October 2025.
Banks already hold the position of trust needed to remain the financial interface in the age of AI. What's missing is the conversational experience consumers now expect, delivered in a way that protects their financial data, and the bank's own IP.
to offer gen-AI financial services
A three-to-one trust advantage that banks need to capitalize on.
Source: McKinsey Global Banking Annual Review survey, May 2025 (n = 30,021).
Your app now reasons across each customer's entire financial life, including data from 3rd-party sources. No new app or assistant to launch, no customer migration and no re-platforming of core systems.
Swipe to see the full architecture.
Data from anywhere lands in one per-customer store, without re-platforming the bank's existing capabilities. Protection is enforced by the architecture itself, not by contract terms with an AI vendor.
A customer can ask anything about PDFs, documents or images, and your app reasons across the whole thing. When an external model is needed, PII is redacted and personal details are disguised, so the AI provider doesn't see customer data or bank IP.
Link outside accounts and everything resolves into one net-worth number, all held in the customer's private vault. What customers bring in stays theirs: your bank reasons across it on their behalf, and it's shared only when they choose.
A real homebuying answer, modeled from income, savings and debt, that remembers the earlier turns as the conversation goes on. Only the figures a question needs are used, and the next step is a real quote from your bank.
Some customers will use ChatGPT, or the next Muse, whatever you do. When your bank is the data steward they pick, your bank stays in the loop. Only the data needed to answer a question is shared, and any action on the customer's account runs through your bank, not an aggregator.
The UI is up to your bank. The same Inrupt capabilities are always at work underneath.
A private intelligence identifies and helps customers connect the accounts, files and feeds that matter.
One data store per customer, ring-fenced from core systems but fed by them, and by new external sources of rich data the customer brings.
Where the bank chooses to surface it, the customer sees exactly what's about to be shared and consents to its use for that purpose.
Only the data points needed to answer a specific question are sent as context to AI models, with identifiers removed and values disguised.
A real answer grounded in the customer's own data. Chat history stays in the vault.
The bank that positions itself as the trusted steward of its customers' data and financial identity takes a position AI platforms can't challenge.
