Erika Sophia Grossbard – We all invest in ourselves in different ways, but the one fundamental human good our nation should invest in is healthcare. To achieve this, the healthcare industry must promote safe innovation in new technology and be legally permitted to do so. Despite the many technological advancements of AI into standard processes across industries, the lack of federal regulation for AI in healthcare has made it difficult to predict how far AI can safely integrate itself into the field.
Given the new normal of AI in healthcare, policymakers are seeking to regulate AI in the practice of medicine. There is a current lack of transparency within the policy framework that deters innovation due to reluctant adopters who have safety and privacy concerns about new software. However, 98% of health care organizations are now in varying stages of considering and adopting AI, embodying the many AI early adopters who found themselves at the forefront, eager to embrace the technology. For AI in healthcare to prosper, it is imperative that all stakeholders trust the technology that will ultimately provide medical care to patients—from regulation to development to integration. Moving forward, the opportunity to regulate can remove much of the uncertainty and allow more societal trust, which is a prerequisite to greater AI adoption among industries.
Accordingly, 2024 promises regulation for AI in healthcare. In recognition of the AI movement, the executive branch finally responded. On October 30, 2023, President Joe Biden enacted an “Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence.” Although the Executive Order only begins to address the regulatory gaps regarding AI in healthcare, it provides some desired guidance for regulatory oversight that will aid in revolutionizing the industry.
Creating guardrails for AI in healthcare can enable groundbreaking approaches to medicine because successful progress may ultimately depend on safer, uniform frameworks. AI will provide virtually limitless applications to address healthcare’s more pressing burdens. Generative AI and precision medicine are two of the most promising trends for AI in healthcare. Generative AI is a branch of artificial intelligence and subset of deep learning that generates high-quality outputs based on training data. Generative AI reduces excessive time spent on mundane yet essential administrative tasks, which grants medical providers more time to spend with their patients. This technology applied to burdensome tasks may also address the problem of physician burnout because generative AI is one of the most powerful productivity tools in generations. Medical providers will then have the bandwidth to expend more resources into prioritizing and achieving the quadruple aim of healthcare: enhancing patient experience, improving population health, reducing costs, and improving provider satisfaction.
In the pursuit of personalized, predictive, and preventative healthcare, implementing AI in precision medicine allows that overdue reality to be taken a few steps further. Precision medicine recognizes that each patient is unique. Although all humans are 99.9% genetically identical, healthcare is not a one-size-fits all approach. Precision medicine requires one of the more advanced applications of AI in medicine because it incorporates genomics, which requires analyzing the patient’s DNA to treat the patient based on genetic uniqueness. Incorporating AI into the genomics market resolves the challenges of acquiring and analyzing near-infinite amount of data, which is a key component to satisfying the demand for personalized medicine.
With the promise of a more innovative and personalized approach to medicine comes the unavoidable risks, including but not limited to issues with uncertainty, privacy, and bias. Although AI algorithms are only as capable as the information inputted by developers permits, AI systems are often deployed without a comprehensive understanding of their expected performance. Consequently, developers ultimately cannot control if AI will generate misinformation that medical professionals rely on to treat patients. Regarding privacy, patients have concerns about their medical records given the many layers of personal health information embedded in them. The digitization of medical records raises concerns of information being acquired by third parties through data being breached or sold. Lastly, AI is manufactured by humans and humans contain biases. There are serious concerns over implicit bias being embedded into the system because algorithmic bias can perpetuate inequalities in healthcare.
Regrettably, nothing—not even extensive testing and training—can prepare developers for the unforeseeable situations an AI system will face in the real world, especially as AI medical tools become more autonomous. Even without laws in place right now that regulate the dynamic nature of AI, developers should hold themselves accountable to ensure their systems can be safely used in medical institutions to assist diverse populations and scenarios. If this responsibility is neglected, a patient may receive improper medical care that could potentially be fatal. Not only is this tragic, but it can also undermine trust in the positive impact of AI in healthcare.
The AI healthcare revolution is already here. Despite the lag in regulations, the industry will continue to innovate in this almost lawless territory until a policy framework is set forth. For now, the implementation of AI in healthcare, specifically through the predicted trends of generative AI and precision medicine, allow for novel opportunities in healthcare. Despite the many concerns of AI, some of which are fundamentally unavoidable, AI’s advancements in healthcare allow the field to become more attractive to investors who are funding AI’s disruptive potential and allowing the progressive AI trends to further develop in 2024.
