Updated: Feb 23, 2022
Artificial intelligence has been defined as the science and engineering of machines and computer programs that more efficiently complete the tasks of human intelligence; it enables problem solving analyses and decision-making capabilities with the use of robust data sets (By: IBM Cloud Education, 2020). A human approach to this problem solving goes as followed: systems that think like humans and systems that act like humans. The ideal approach involves utilizing artificial intelligence, encompassing systems that think and act rationally (By: IMB Cloud Education, 2020). More work can be completed using the ideal approach of artificial intelligence, as it is the computers thinking and acting as rational agents.
When artificial intelligence is used in the healthcare setting, it is labeled as cognitive computing (CC). For instance, IBM Watson for Oncology has allowed physicians to be able to explore various treatment options for patients by using computer-based algorithms of hypothetical and real patient medical records (Gerke, et al., 2020). Outcomes from millions of patients are grouped together in a regression analysis that indicates the best method of treatment for said individual. This data-driven analysis in healthcare holds the capacity to increase knowledge and efficiency. Unfortunately, there are ethical concerns and implications of the use of artificial intelligence in healthcare. A few of the ethical challenges include informed consent to use, safety and transparency, algorithmic fairness and biases, and data privacy (Gerke, et al., 2020).
Although it may be true that consent forms are utilized for the authorization of patient information, this consent tends to be blurry and vague. This poses the question of just how much responsibility a physician holds to explain the intricate methods of artificial intelligence, the uses of the information, and what data is being inputted into the system. Many times, the physicians themselves do not understand the complexity of the systems. Aside from the physician aspect of consent, there is also artificial intelligence in any health application or chat box. User agreements are vast, and most people do not fully read or understand what they are consenting to (Gerke, et al., 2020). This issue of consent violates the Health Insurance Portability and Accountability act of 1996, commonly referred to as HIPAA, which was enacted to protect patient health information.
Similarly to the issue of consent, there is also an ethical concern regarding the safety and transparency of artificial intelligence. If a data set is flawed, poor medical treatment could be administered as an effect. If the data is reliable and accurate, the artificial intelligence system will perform better, but these algorithms need to be continuously refined. As for transparency, it would be ideal for all the data and algorithms to be available to the public, but this has adverse effects on intellectual property and the potential for cybersecurity risks (Gerke, et al., 2020).
Another key point of concern comes from the biases that occur in A.I. algorithms. When phenotypes and genotypes are entered into these algorithms, biases grow and diagnoses and treatments could be rendered ineffective (Gerke, et al., 2020). Put simply, human generated algorithms are only as fair and efficient as is the input data.
The final ethical concern of the use of A.I. in healthcare encompasses them all: data privacy. This privacy is a fundamental human right. Innovation does not need to encroach on this human right. The question of data ownership has been posed by many individuals, as they report being nervous about where their information is going and who has access to it. Health data has a high monetary value and has been sold in many instances (Gerke, et al., 2020). If patient data is leaked or sold, discrimination has the potential to occur.
Ultimately, the use of A.I. in healthcare settings has proved to foster innovation, but not without it being at the expense of millions. It is imperative to evolve higher safety and transparency techniques so that individuals can feel confident about their health data not being breached or sold.
1. By: IBM Cloud Education. (2020, June 3). What is Artificial Intelligence (AI)? IBM. Retrieved February 16, 2022, from https://www.ibm.com/cloud/learn/what-is-artificial-intelligence.
2. Gerke, S., Minssen, T., & Cohen, G. (2020). Ethical and legal challenges of artificial intelligence-driven healthcare. Artificial Intelligence in Healthcare, 295–336. https://doi.org/10.1016/B978-0-12-818438-7.00012-5.
Author: Sarah Ellis
Editor: Lauryn Agron
Health scientist: Sarah Ellis