BILL has launched agentic AI capabilities that flag unusual invoices, centralise AI controls, and let finance teams query data in chat.
US-based BILL, a finance platform used by nearly 500,000 businesses, has introduced new agentic AI capabilities designed to give finance teams more control over how AI operates in their businesses. The features cover payment error detection, a central hub for managing AI, natural language chat, AI-assisted onboarding, and automatic document labelling, and are rolling out to customers now. The launch addresses a trust gap identified in BILL’s own research: in a survey of US accounting firms, nearly 90% described themselves as ambitious about AI, yet only around half said they trust AI data accuracy ‘somewhat’, while almost nine in ten finance teams said they want centralised or in-workflow control of AI.
Risk intelligence and AI controls
In fiscal year 2026, BILL flagged around USD 1.5 billion in customer payments that resulted from simple mistakes or unusual activity, such as a vendor sending an invoice for the wrong amount. Its new risk intelligence capability checks invoices before payments are scheduled and flags those that deviate from a vendor’s established billing history during the approval stage. Between August and September 2026, the feature detected an average of USD 138,000 in mistakes or unusual payments per customer, saving customers more than 1,800 days of manual work in total. A new consolidated hub allows customers to monitor the actions BILL’s AI has taken, switch AI tools on or off, and set how much the AI does on their behalf.
René Lacerte, CEO and Founder of BILL, stated that finance leaders should be able to decide how and where AI works for them, and that the new capabilities give them the choice to use AI to protect their money and log AI actions. According to Lacerte, BILL’s AI is built on more than USD 2 trillion in payment volume, giving it exposure to a wide range of payment, vendor, anomaly, and fraud patterns.
Chat, assistant integrations, and onboarding
Finance teams can now ask BILL questions in plain language through a chat feature, currently in beta, covering topics such as which vendors need paying, where an invoice stands, what is driving card spend, and how month-end close is progressing. BILL notes that AI-handled support conversations grew more than 250% in a year. Through an external Model Context Protocol (MCP) server, customers and the more than 9 million members of BILL’s network can also connect their accounts to AI assistants and query their BILL data from there, as developer use of BILL’s APIs has grown sevenfold in two years.
BILL’s AI-assisted onboarding personalises the business profile, recommends products, and helps users add vendors and assign default approvers, cutting platform setup time by up to 50% and vendor customisation time by up to 70%, with users reviewing all changes before they apply. New inbox capabilities also identify and label incoming invoices and W-9 forms automatically. By keeping approval and oversight with finance teams, BILL is positioning its AI features to address the hesitation SMBs have about fully trusting automated financial decisions.
