- Key insight: Intuit has added Meta's Muse to its AI suite.
- What's at stake: Payment companies are pursuing agentic AI as a competitive differentiator.
- Expert quote: "As these tools scale, access to high quality data is essential for both the industry and regulators."—Kaiko's Thomas Probst
Intuit is the latest small business services provider to embrace the social network's AI assistant, adding support for Muse within the QuickBooks connector. "We know small businesses have been leveraging AI agents like Muse and others in the market to help streamline tasks and get key insights on their business that help them understand their performance and know where they stand financially," Alex Balazs, executive vice president and chief technology officer at Intuit, told American Banker.
Intuit is not alone in its vote of confidence in Muse. While Amazon has engaged in a high-profile
How Intuit's using Muse
QuickBooks customers can use Muse to help manage cash flow and payments, and streamline more complex parts of operating a business like running payroll, closing books or filing taxes.
Balazs said Intuit's existing tools back Meta's AI experience, which he says includes
permissioned data and Intuit's existing experience with small business operations and compliance.
Just as search engines and app marketplaces created new paths to trusted products, AI agents like Muse can help connect people to the right tools and platforms for what they want to accomplish, Intuit said, adding it has also partnered with other AI companies like OpenAI, Anthropic, Google and Perplexity.
At Intuit, Muse is part of a growing stack of small business technology and AI that is designed to augment its encroachment on banking. Intuit has long sold accounting technology to small businesses, and in recent years has
Earlier Intuit AI initiatives include making QuickBooks and Mailchimp available directly inside Claude, ChatGPT and Perplexity. This enables businesses to manage invoicing, get answers to payroll questions, and follow up on cash flow, with QuickBooks Capital's lending available in Claude and ChatGPT.
Balazs said Intuit's goal is to provide actionable information rather than generic digital answers, using large language models as "reasoning engines" that aid work but don't replace it. The firm's argument is it has more than 10 million businesses across 307 industries that can provide usable expertise for the AI models.
"These AI agents provide an intuitive entry point to ask questions and get oriented, and serve as a new front door to help businesses connect to the right tools that actually power the work," Balazs said.
On the frontier
Similar to Intuit, the payment firms that are adopting Muse are using it mostly to add more heft to existing AI programs.
In addition to enabling Must to perform searches and checkout, PayPal also supports ChatGPT Checkout, which lets users complete purchases using a PayPal wallet. PayPal also processes payments for merchants using OpenAI's Instant Checkout. PayPal's AI suite additionally includes Agentic Commerce Services, which sells payment support, order management and connections between merchants and product data, fulfillment and AI-powered checkout.
Another feature called PayPal "agent ready" enables merchants to accept payments via different AI programs; while another feature called "store sync" makes merchant data discoverable within AI channels, including the ability to route orders to their existing fulfillment and management systems.
At Shopify, Muse scours Shopify's catalog to find the best fit for a shopper's AI prompt and then supports checkout through the Universal Commerce Protocol, which guides AI-powered shopping and checkout. This means Shopify merchants are discoverable and purchasable in Muse by default, with product data flowing through Shopify Catalog, according to Shopify.
Google and Shopify co-developed the Universal Commerce Protocol, which is a set of standards that are interoperable with AI from Perplexity, Gemini, ChatGPT and other products. The standards are designed to vet agents, to ensure intent and make it easier for AI agents to communicate.
While Muse and other similar "frontier" AI assistants have a wide range of uses, there are real opportunities for payment firms, which are already building the infrastructure needed for transactions initiated by AI agents on behalf of their users, according to Thomas Probst, a research analyst at Kaiko.
"Stablecoins are particularly well suited to this shift, especially for small and frequent payments between machines, as they are programmable, settle within seconds at any time and work across borders," Kaiko's Probst told American Banker. "Firms that build trusted infrastructure around them could capture a lot of value."
Read more:
As fintechs nab bank charters, what's happening to BaaS? Overheated data center knocked 23 credit unions offline Chris Britt still says banking is broken, but Chime can fix it Credit unions bulk up a business that banks long dominated
On the risk side, one of the main challenges is the quality of the data these agents rely on, according to Probst.
"Prices, rates and market information need to be accurate, reliable and traceable, because an error in the underlying data can spread across a vast number of decisions at once, with increasingly limited human oversight," Probst said. "As these tools scale, access to high quality data is essential for both the industry and regulators."
