Healthcare’s Red and Blue Pill: AI

Kirsten Brueggemann

Associate Editor

Loyola University Chicago School of Law, JD 2025


Artificial Intelligence (AI) has gained widespread attention, often perceived as a buzzword. Recently, concerns about its potential dangers and issues with plagiarism have surfaced. However, AI holds immense promise for transforming industries reliant on data analysis and predictive algorithms, especially in healthcare. AI can significantly improve healthcare by aiding in diagnosis, optimizing patient outcomes, reducing costs, and saving time.

What is AI?

AI combines many areas of technology: machine learning—a field dedicated to the study of algorithms that allows computer programs to improve automatically through experience. Reinforcement learning, another facet, involves technology learning by trial and error, or by expert demonstration. Additionally, deep learning is a subset of algorithms that learns by using large collections of processes and exposing them to a vast set of examples.

AI uses computers and machine processes to perform complex automatic functions and to simulate human intelligence. AI can help providers identify patterns, trends, and anomalies. This allows providers to predict patient outcomes with more accuracy and analyze different treatment options according to individual patient data.

The AI cure?

AI can be used in the healthcare industry to improve diagnosis, treatment, reduce costs, and save time for patients as well as providers. AI can aid in improving diagnosis, even for physical conditions that have no obvious symptoms. As a current example of the benefits of AI in healthcare, the Mayo Clinic is using AI software to detect and diagnose serious or complex heart problems. Furthermore, a Mayo Clinic study applied AI towards a screening tool for people with a heart problem that has no obvious symptoms, also known as left ventricular dysfunction. In the study, the AI screening tool found people at risk of this condition 93% of the time. For reference, a mammogram is correct about 85% of the time. This is one area where AI has been implemented and shows significant promise. AI screening technology along with other data collected from other wearable devices such as sensors, can help medical providers diagnose and treat patients more accurately.  AI can also aid in optimizing treatment plans according to patient data and has the ability to make medication and treatment recommendations according to a patient’s medical history and lifestyle.

AI also has the potential to lower health care costs by reducing administrative burdens and streamlining provider’s administrative responsibilities. Administrative burdens, such as patient record-keeping, prescription management, and insurance building, account for approximately $1 trillion dollars in annual healthcare costs within the United States. By minimizing administrative burdens for providers, this enables them to bring down their costs. Providers can also streamline their responsibilities with AI by using AI-powered scribing solutions to take detailed notes of provider-patient interactions to better capture the visit. They can also use AI to process prior authorizations, that takes up a significant portion of provider’s time and cost.

The implementation of this system empowers healthcare providers to allocate a greater portion of their time towards direct patient care, as opposed to being bogged down by onerous administrative duties that diminish their capacity to interact with patients effectively.

The thorn in AI’s side

AI relies on access to large sets of data and its future depends on medical providers being able to access and share this information with each other and patients. One significant obstacle AI faces is patient privacy. AI companies in the healthcare space need to adhere to Federal and state laws regulating individual data privacy.

Because the collection and use of patient information handles sensitive individual data, the Health Insurance Portability and Accountability Act (HIPAA) may be implicated. One method proposed is de-identification, where certain data points of an individual’s information are removed prior to uploading to AI software. De-identification of data depends on the data being analyzed and the laws and regulations that are implicated. For example, if HIPAA is implicated, de-identification of protected health information requires the removal of certain identifiers via the Safe Harbor method or an expert determination that data is considered de-identified. Certain Safe Harbor identifiers include names, email addresses, and dates relating directly to an individual.

Despite the removal of certain identifiers, some claim that there is a significant threat of re-identification. Re-identification is when individuals can be re-identified even when the dataset is de-identified beyond the Safe Harbor method required by HIPAA. However, a study conducted by MIT suggests that the risk of patient re-identification via publicly available health data is extremely low. The benefits of sharing de-identified data far outweigh the potential risk to patient privacy. There is harm when data is not shared, especially as AI’s entire selling point relies on the sharing of information between providers and the collecting of data from patients.

Perhaps more significant threats to patient privacy are data breaches. With the adoption of Electronic Health Records and with healthcare entities using Cloud Services and Information Systems, the healthcare industry has become plagued recently by data breaches. Data breaches pose a significant threat to the use of technology in the healthcare industry. However, they are not unique to AI.

AI’s Future in Healthcare: Balancing Progress and Safety

The benefits of AI in healthcare outweigh its potential hazards. AI has the ability to improve patient outcomes, improve quality and course of treatment, reduce the cost of healthcare, and save time for both the provider and the patient. Patient Privacy is a significant obstacle to healthcare providers and AI companies as patient data is an integral part of AI within healthcare. However, healthcare providers should work to balance patient privacy and innovation through AI to ultimately benefit their patients.