AI powered automation – useful tool or slippery slope?

AI Powered Automation

In the world of chatbots, virtual assistants, and predictive modeling, it’s hard to find an industry that AI has not infiltrated. When we consider the addition of AI features into our billing processes, do we really understand the risk?

AI can be useful for performing repetitive and monotonous tasks, but it does not come without risk. Is the dependence on AI resulting in a loss of professional skills? Perhaps. There can be a lack of transparency, consistency, and fear of surveillance and manipulation with AI. While the United States is the global leader in AI capacity and development, there is little to no legal regulation related to AI. So how do we prevent any existential threats from becoming a disruptive force to our operation?

AI powered automation introduced into the billing process often promises increased accuracy in the billing process while abolishing manual input. Is that a great idea? Maybe. Maybe not. There is no argument that the integration of AI saves time and money. In an industry where accuracy is critically important, how do we determine if we have gone too far?

Have you ever watched a movie with captions on and noticed the words on the screen don’t always match what is being said? Sometimes a misunderstanding can result in a hilarious interpretation floating on the screen. AI cannot always understand accents, slang, or regional dialect and routinely gets it wrong.

The billing process is built on accuracy and a comprehensive understanding of the rules and regulations we are required to abide by. In complicated scenarios where we must weigh the facts against the regulations, are we going to trust a computer to make the right decision? What if it doesn’t? Can AI correctly determine if a GM, GY, GA, GZ, QL or some other special modifier should be added? Doubtful.

Does that increase the risk of an audit that may uncover the AI program got it wrong time and time again and you owe a large sum of money back? You bet it does! Did you know CMS is already using AI for fraud detection?

Can a human make the right determination better than a program? There isn’t enough information yet to answer that question definitively. Never underestimate the power of highly skilled people to perform the important functions in the revenue cycle to maintain compliance and accuracy.