AI and automation in medical billing: Separating hype from reality

This post is the first installment of the Medical Billers’ Triple Threat series that explores how billing leaders are navigating AI, compliance, and cybersecurity.

Artificial intelligence (AI) is starting to dominate the revenue cycle — and medical billers are trying to cut through the noise. Many are struggling to break free from manual workflows or hesitancy around AI; 42% of billers still haven’t adopted any automation or AI in their organization, according to Tebra’s 2024 survey of medical billing professionals​.

However, payers aren't waiting around. They're using AI themselves to deny claims in seconds — leaving billers with lost revenue.

While technology vendors pitch their AI solutions as a cure-all, billing professionals need practical tools that bring measurable results: fewer denials, faster reimbursements, and reduced rework time. 

For billing companies navigating today's technology landscape, understanding the practical differences between automation and AI is pivotal for making smart investments that deliver long-term value. We asked revenue cycle management experts to separate what's hype from what's real — and how to navigate what's next.

If you’ve been evaluating new billing technology, you’ve likely seen the terms “automation” and “AI” used interchangeably. But they serve different purposes.

Automation executes repeatable, rules-based tasks according to predefined parameters:

AI can learn patterns, make predictions, and adapt its approach:

Alexis Marshall, Client Solutions Manager at Medical Billing Strategies, describes how AI goes further than automation: "[AI] can learn fee schedules and reimbursement rates. That’s where it becomes really powerful. In my experience, automated systems don’t catch discrepancies. Where automation is a 1:1 match, AI brings intelligence. It can detect underpayments. Automation can’t do that.” 

However, AI can still create errors — so it's important to continuously monitor outputs. Loren Dilger, CEO of reCLAIM Billing Solutions, describes the risk of not doing so: “Misunderstanding AI is the biggest risk I see. Companies are trusting tools they don’t understand — and skipping the oversight they’d never skip with a human.”

Before investing in AI-powered solutions, billing companies need to cut through buzzwords and assess real capabilities. Here are the questions to ask vendors.

How is the AI trained, and who reviews its decisions? 

Ask about training data, quality controls, and human oversight processes.

Dilger emphasizes human oversight: "You wouldn't put an employee on a task without a manager in the loop. Why do we do that with AI?"

What measurable outcomes can you tie to payer performance? 

Request specific metrics that show improvements in areas such as clean claims rates, denial reductions, or days in A/R.

Jeff Hillam, CEO of Red House Medical Billing, recommends requesting meaningful data from vendors: “If a vendor tells me they ‘do AI,’ I immediately ask them to define it. Often, they’re not offering any meaningful data about how AI improves outcomes with payers, only how it makes processes easier for us. [AI vendors] spend so much time just learning how to bill... they’re tech folks, not RCM folks.” 

What controls exist for HIPAA compliance and error prevention? 

Ensure the vendor maintains robust data security and complies with HIPAA.

Dilger notes an essential step: "Make sure you get a business associate agreement (BAA) in place. If you're working with a vendor that doesn't first send you a BAA, you're not working with the right vendor."

It’s crucial that not only vendors but also billers comply with HIPAA. Aimee Heckman, Director of Revenue Cycle Management at Ash Business Solutions, cautions billers who use public AI tools: "A word of caution for billers who want to dabble in AI using some of the readily available tools like ChatGPT, Grok, and Google Gemini: be sure to follow HIPAA guidelines regarding PHI and only use de-identified data if you intend to put claims data into one of these tools."

Dilger, who is an AI platform builder within the billing space, offers the following tips for working with a large language model (LLM).

He also emphasizes the value of implementing AI workflows: "We're not replacing people — we're trying to get the mundane out of their day.” 

Medical billing companies face increasing pressure to adopt new technologies while maintaining operational stability. The key is to move at a strategic pace and leverage automation now for immediate gains while thoughtfully exploring AI capabilities for the future.

By understanding the real-world applications of both, billing companies can turn technology from a source of noise into a competitive advantage — and improve efficiency, reduce costs, and strengthen their financial performance.

Download our complete guide to facing the triple threats in medical billing: AI and automation, cybersecurity risks, and increasing regulatory scrutiny. Get expert strategies from industry leaders who are successfully balancing innovation with practical results.