AI expense management at Expensify just got more hands-off. On July 1, 2026, Expensify expanded its Concierge AI to handle far more of the expense workflow. Customers can now configure accounts, automate tasks, and analyze spending in plain English. The tools work by email, text, or inside the Expensify app.
The goal is less manual work and clearer spend visibility. Instead of clicking through menus, users just ask. Concierge understands the request, does the work, and explains what it did.
AI Expense Management: What Concierge AI Can Now Do
The update adds five capabilities to Expensify’s AI expense management. Conversational analytics lets users ask about spending by employee, category, merchant, workspace, or period. Concierge replies with explanations and visual charts. Chat-based management lets users create, edit, categorize, tag, submit, approve, and reject expenses by command.
Workspace administration goes further. Admins can set policies, create rules, update settings, and troubleshoot issues just by asking. Cross-channel access means the same AI works over email, text, or the app. The fifth piece is the boldest. Autonomous workflow agents, now in beta, follow custom instructions to run routine tasks on their own, with a full audit trail.
Access is immediate. Customers email Concierge, reply to a Concierge email, text 47777, or use the app. The system learns and improves with each interaction. Expensify CEO David Barrett framed the point simply: “AI should eliminate work, not create another interface to learn.”
Why the AI Expense Management Push Matters
The launch fits a broader shift in finance tools. Software is moving from menus and forms toward conversational, agent-driven systems. Expensify has leaned into that. Last month it launched an MCP integration so outside AI assistants, including ChatGPT and Claude, could query Expensify data directly. This update brings that same capability inside the product.
It also sharpens competition. Spend-management rivals like Ramp have pushed AI agents into procurement and expense workflows. Expensify’s answer is to make its own agent reachable from a text message.
This is not Expensify’s first agent move. It rebuilt Concierge into a contextual expense agent in late 2025, blending AI with around-the-clock human support. Barrett has argued the future lies in collaboration between humans and machines, not full autonomy. The July update leans harder into automation while keeping that audit trail and human backstop.
For finance teams, the pitch is concrete. Less busywork, faster answers, and routine approvals handled without opening a dashboard. Expensify, used by 15 million members worldwide, is betting that plain-English control becomes the default in AI expense management.
