AI vs AI
Why health insurance carriers use AI and healthcare providers don’t....... until now
Dalton Han - Co-Founder - Red Sky Health
Healthcare Providers Have Been Slow to Adopt AI for Claims Processing
Large insurance carriers have been leveraging artificial intelligence (AI) across their operations for years. From automating claims intake and adjudication to detecting fraud and enhancing customer service, AI technologies such as Natural Language Processing (NLP), robotic process automation (RPA), and predictive analytics have become deeply integrated into their systems. These tools have allowed insurers to improve efficiency, reduce costs, and gain powerful insights from data. Yet despite these advances, healthcare providers, the other side of the healthcare reimbursement ecosystem, have been much slower to adopt AI in their claims processing. This raises an important question: Why?
Historically, AI had high barriers to entry. AI solutions required substantial upfront investment in technology, infrastructure, and skilled personnel. Large insurance carriers have the financial resources and scale to absorb these costs to justify the long-term return on investment. By contrast, healthcare providers, particularly small and mid-sized practices, often lack the capital or technical expertise to invest in such technologies. Implementing AI requires not just software, but also data integration, staff training, and ongoing system maintenance—obstacles that can be prohibitive for many providers.
The structure of the healthcare provider landscape also contributes to this lag. Unlike the insurance industry, which is dominated by a few large carriers, the provider ecosystem is highly fragmented. There are thousands of hospitals, clinics, and private practices across the country, all using a wide array of electronic health record (EHR) systems and billing platforms. This fragmentation makes it difficult to deploy standardized AI solutions at scale. What works for one provider may not be compatible with another’s systems or workflows. This lack of standardization slows innovation and increases the complexity of integrating AI into everyday operations
The relationship between healthcare providers and insurers further impedes AI adoption. In the current healthcare system, insurance carriers dictate the rules of the game. They establish coverage guidelines, determine coding requirements, and control payment timelines. This power imbalance creates a dynamic like a Las Vegas casino: providers must play by the house rules, with limited ability to influence or change the game. Even if a provider implements advanced AI tools to streamline claims processing, they are still subject to the often opaque and shifting policies of the insurers. This leaves little incentive for providers to invest in technologies like AI that may not deliver a return under the current constraints.
In addition, while insurers benefit from AI by analyzing massive datasets to predict claim validity or identify fraudulent activity, providers typically do not have access to the same breadth of information. Without a full view of payer behavior and claim outcomes, providers are unable to train AI models as effectively. They are essentially operating with incomplete information, limiting the potential power of AI to improve their processes.
The Rise of Everyday AI in Health Insurance
The advent of large language models (LLMs), such as GPT and Claude, has dramatically changed this landscape. Today, AI is not only easier to use but also more effective, offering healthcare providers powerful new ways to manage administrative burdens like insurance claim denials.
These models are pre-trained on massive amounts of data and can understand and generate human-like text. This means healthcare organizations no longer need to train models from scratch. Instead, they can leverage existing models that already "know" how to process, interpret, and interact with medical and insurance-related language. The result is faster implementation, lower costs, and more accurate outcomes. LLMs are also continuously improving. Advancements in machine learning and natural language processing, newer models become smarter, more reliable, and more adaptable with each release. These improvements enable AI tools to handle increasingly complex tasks, such as identifying errors in claim documentation, predicting claim outcomes, and even generating appeal letters automatically.
Another factor driving this change is the rapid growth of the AI workforce. There are now significantly more AI engineers, data scientists, and machine learning specialists than ever before. This expanding talent pool makes it easier for healthcare providers to find the expertise they need to integrate and customize AI solutions. In parallel, the rise of no-code and low-code platforms means that even non-technical staff can deploy AI tools with minimal training.
Most importantly, society as a whole has become more familiar and comfortable with AI. What was once viewed as a futuristic concept is now part of daily life. From virtual assistants to AI-powered chatbots, people interact with AI regularly—at work, at home, and even in healthcare. This increased comfort level makes the transition to AI-driven processes, like insurance claim denial remediation, smoother and more widely accepted by both staff and patients.
In summary, AI was once a challenging, costly, and inaccessible tool for healthcare providers. But thanks to the rise of LLMs, the growing AI workforce, and the normalization of AI in everyday life, healthcare organizations can now leverage this technology with greater ease and efficiency. This transformation marks a significant step forward in ultimately improving patient care.
