A billing write-off in legal practice is a reduction applied to billed time or fees before or after invoicing a client—either because time was recorded but deemed non-billable (a write-down before billing) or because a client disputes and the firm reduces the invoice (a write-off after billing). Write-offs are a normal feature of legal billing, reflecting the gap between time recorded and time considered reasonable, necessary, and defensible to charge for in context.
AI changes the billing write-off calculation in a fundamental way. When a task that previously required four hours of associate time can now be completed in forty minutes with AI assistance, the firm faces a structural question: should it bill four hours at the pre-AI rate (which may feel dishonest or invite dispute), forty minutes at the pre-AI rate (which sharply reduces revenue), forty minutes at a higher "AI-enhanced" rate, or transition the matter to a flat fee that captures value independent of time? The answer varies by firm philosophy, client relationship, and billing arrangement—but AI has made the question unavoidable.
Write-off analytics have become more sophisticated in AI-enabled practice management platforms. Firms can track write-off rates by attorney, practice group, client, matter type, and time period, and increasingly by whether AI tools were used on specific tasks. This data reveals patterns—systematically high write-offs on AI-assisted research, for example—that help firms calibrate billing practices and pricing strategy for AI-augmented work.
AI-driven efficiency creates billing model tension in hourly-rate practice. When clients know that AI can compress research and drafting timelines, they increasingly question invoices for work that an AI could perform in a fraction of the recorded time. Outside counsel guidelines from institutional clients are beginning to address AI billing directly, with some prohibiting billing for AI tool costs separately and others requiring disclosure of AI use on billed matters.
Write-off rate is also a key profitability metric. Firms with systematically high write-offs—billing significant time that clients dispute or that attorneys self-write-off before billing—are losing revenue and obscuring true matter economics. AI adoption, if managed without billing model adjustment, can increase write-offs as attorneys record time on tasks that AI completes faster than the recorded hours suggest is plausible.
The constructive response is to use AI billing data to inform pricing model evolution. Firms that track AI-assisted task efficiency and use that data to set flat fees, subscription pricing, or AI-adjusted hourly rates for specific services can maintain profitability while offering clients fee structures that reflect AI-era economics. Write-off analysis is one input to this calibration.
Practice management platforms like Clio, MyCase, and Filevine provide billing analytics that can be segmented to identify write-off patterns. Some integrations with AI tools allow time entries to be tagged when AI was used, creating data sets that reveal how AI adoption affects billing outcomes by task type. This data supports evidence-based decisions about pricing, billing policy, and write-off thresholds.
AI-assisted billing review tools—a nascent category—analyze time entries for anomalies, suggest write-offs for entries that appear excessive or duplicative relative to AI-enabled task norms, and flag billing practices that are likely to attract client disputes. These tools are primarily marketed to corporate legal departments evaluating outside counsel invoices but are beginning to appear as firm-side self-audit capabilities as well.
The deeper transformation AI drives in billing is toward value-based models that reduce or eliminate the write-off problem by decoupling fees from recorded time. When a firm prices a matter based on value delivered rather than hours worked, write-offs become less relevant—the question is whether the flat fee is adequate for the work, not whether the hours recorded are justifiable.