By Healthesystems
These days, it seems like Artificial Intelligence (AI) is everywhere. It is revolutionizing how we do just about everything in our lives – and that includes how we access healthcare services. If you’re covered by a health plan, you’ve likely seen your insurer’s website or app incorporate virtual assistants which help you to find doctors, estimate the cost of care, and schedule appointments.
What may be less obvious to the average consumer are the numerous applications of AI behind the scenes to improve healthcare delivery from quality, innovation, and efficiency perspectives. From applications in drug discovery, chronic disease management, diagnostics, medication management, and more, healthcare systems are only scratching the surface of AI’s potential.
So how does all of this translate into workers’ comp? What are some of the opportunities and pitfalls to consider, especially when considering the regulation of AI? We spoke with Sandy Shtab, Healthesystems VP of Industry and State Affairs, to gain her perspective.
How is AI being used in workers’ compensation healthcare?
In the realm of workers' compensation, AI programs can be utilized for many functions – from the physician’s office to the mailroom and all the way through to the claim file.
For example, when an injured worker requests a second opinion or an adjuster requests an Independent Medical Examination (IME), physicians often have to comb through years of treatment records, which is time consuming and can be frustrating if there are duplicate records or conflicting information. For physicians, today’s AI technologies can:
- Chronologically order a patient’s medical record
- Summarize and easily identify if there are inconsistencies in the records that may have been overlooked by the previous provider
In terms of claims organizations, there is a plethora of technologies that provide mail room support, eliminating manual and repetitive tasks previously performed by administrative staff. These services can:
- Sort and distribute inbound mail to the correct claims professional
- Load structured and unstructured data from electronic health records into claim files and attach digital images of MRIs and other diagnostics
- Quickly route medical bills to bill review entities
Additional AI tools in workers’ comp include:
- Decision support tools to help the claims professional take in more information, enabling them to speed up care for the injured worker
- Systems to analyze large volumes of medical records and claim data and identify potential patterns of fraud, waste, and abuse
- Predictive analytics tools to enable adjusters to forecast disability durations and medical reserves
How can AI benefit injured workers?
AI can benefit injured workers by ensuring they receive more personalized care tailored to their specific needs. It can offer helpful functions like sending the injured worker appointment reminders via text or in-app notifications when it is time to take their medication. It can even help the patient record when they have taken their medication.
There are also some cool tools that enable the provider to do remote patient monitoring, making it easier to keep track of vital signs and symptoms. These are the kinds of details that are important for the physician to know but might be challenging for the injured worker to recall days or weeks later at their medical appointment. These tools can help ease the burden for an injured worker.
What are the pitfalls or risks of using AI in healthcare?
While we have just scratched the surface of the potential for AI in workers’ comp and healthcare, there are many who are rightfully concerned about the potential risks. One of the biggest concerns is the accuracy of these tools and transparency to the user in how the AI algorithms work.
There are concerns about biases and accuracy within the systems, which could lead to missed diagnoses or incorrect treatment recommendations. And it goes without saying that data privacy is a huge concern, as AI systems often require access to electronic health records and sensitive patient information.
Are there any ethical or data privacy issues?
Absolutely! Ethics and data privacy issues should be a concern for all of us. Effectively training AI programs requires those systems to access vast amounts of patient data, raising concerns about data security and privacy. There must be precise measures that protect patient information from unauthorized access and ensure compliance with data protection regulations.
There are also ethical considerations, such as ensuring that AI systems are used fairly and do not discriminate against certain patient groups. For example, if AI systems had access to employer health plan data or patient information and could crosswalk that data to public databases like state workers’ compensation systems, they might be able to more accurately predict risk and claim costs – but they may also use the information to their competitive advantage.
Easy access to litigation or settlement information may sway an employer or carrier away from hiring a qualified employee. There are active investigations into how health plans are using AI to deny authorization of services. Easy access to those types of information could be used to discriminate against individuals and entities.
How are federal and state agencies attempting to regulate AI in healthcare?
The new administration has taken a different approach than its predecessor, which was focused on data privacy protection and eliminating bias in AI systems. In recent months, there have been orders revoking prior policy. These represent a departure from consumer and data protection concerns and instead favor the elimination of any regulatory obstacles to AI innovation.
There is also a recent directive from the White House to incorporate AI into classroom settings, which will certainly have an impact down the line on AI in healthcare. Congress has not taken any meaningful action as of this writing. However, there are many proposed bills that would incorporate AI into improving access for patients to AI-enabled clinical technologies, speeding the FDA approval process and FDA-cleared devices and improving patient outcomes.
In recent years, the states have also been active in considering legislation about AI, much of which is focused on consumer disclosure requirements by healthcare entities that are using AI for decision support or clinical functions. However, in May 2025, an amendment was proposed to the new Congressional budget bill that would ban states from regulating AI for the next 10 years. If the bill passes with this stipulation, it will block state efforts and halt future state regulation of AI in healthcare.
Critics argue that this would leave consumers without protection on how AI uses their data or makes decisions. And with no comprehensive federal framework currently in place, this would impact how organizations approach compliance, operational workflows, and safeguards related to AI-driven processes.
Any last thoughts on AI and the future of workers’ compensation claims?
Workers’ comp is a business built around helping people to receive medical care at a reasonable cost when a worker is injured on the job. While AI can help support injured workers and automate tasks for claims and medical professionals, what it cannot do is replace the human element.
The human aspect of claims handling will become increasingly important and more valuable as we begin to rely more on AI tools. AI will certainly get better at replicating human behaviors, but the jury is out on whether something like an app can demonstrate empathy or compassion in a way that feels authentic. I don’t think the human element is something AI can ever replace.
Sandy Shtab is VP of Industry and State Affairs at Healthesystems. She leads the Advocacy & Compliance team, which is essential in the management of pharmacy and ancillary benefits for Healthesystems clients. The team regularly tracks regulatory activity via the Regulatory Recap e-newsletter, which is also available on the Advocacy & Compliance page of the Healthesystems website.
