AI Resilience Report forBilling and Posting Clerks

If you're a young person eyeing office jobs, here's the honest picture: billing and posting work is one of the areas where AI is moving fastest — but there's still a role for humans who can think, catch mistakes, and talk to customers. The U.S. Bureau of Labor Statistics ^[1] groups billing and posting clerks with occupations where demand is expected to be limited for billing and posting clerks; procurement clerks; credit authorizers, checkers, and clerks; customer service representatives; and nonmedical secretaries and administrative assistants, because AI integration into workflows is expanding. Concretely, BLS projects billing and posting clerk employment to slip from about 429,800 to 427,900 (a ~0.4% decline) between 2024 and 2034 ^[1] — not a cliff, but flat in a growing economy.

Much of this comes from tools that read invoices, match them to purchase orders, and post payments automatically. In healthcare billing, HFMA reports that 27% of organizations are actively deploying AI at scale ^[2] across multiple functions, and 53% are conducting pilots, with agentic AI tackling accounts receivable follow-up, underpayment management, denials management and cash posting — labor-intensive, rules-governed tasks. On the corporate side, an IFOL/SAP Concur 2026 study ^[3] found 19% of organizations already use AI within AP and a further 30% plan to adopt within the next 12 months.

That said, humans still shine at the "gray area" work — verifying odd charges, chasing customers about disputes, and handling exceptions.

Adoption is accelerating because the tools are cheap, off-the-shelf, and the ROI is obvious. McKinsey observes that finance teams are moving beyond experiments into scaled deployments ^[4], and HFMA notes McKinsey estimates AI in the revenue cycle could lead to a 30% to 60% reduction in cost to collect ^[2]. Billing tasks are also very "AI-friendly": they're digital, rule-based, and repeat millions of times, which makes them ideal for intelligent process automation that combines RPA with AI ^[5].

Adoption headwinds do exist, though. Brookings researchers warn that clerking and administrative occupations show low adaptive capacity to disruption ^[6], meaning workers in these roles may struggle to pivot without support, which is pushing employers to redesign roles rather than simply cut them. HFMA's survey found just slightly more than half of healthcare finance leaders describe their teams as "somewhat prepared" (44%) or "very prepared" (7%) ^[2] for the AI-enabled revenue cycle — so training gaps are slowing things down.

The encouraging news: employers still need people who can handle exceptions, communicate empathetically with confused customers, and audit what the machines do. If you build skills in data verification, customer communication, and basic AI oversight, you'll be far more valuable than a clerk who only keys in numbers.