When providing leadership and direction to accountable care organizations that collaborate to deliver human-centric high-quality, coordinated services and healthcare, ethics are at the forefront of data-driven decision-making. Evolving faster than ever is the blended approach of ethics and artificial intelligence, which is a systematic normative reflection based on holistic, multicultural and assimilating frameworks of interdependent values, morals, principles and actions.
AI and ethics need tight coupling going forward as the adoption rate grows for these capabilities in healthcare management. The most common approach of bringing together these disparate data sources in control of multiple healthcare facilities is by virtue of federation. Whether these federated data sources are consumed at-source or brought into a centralized database, the ethical standards and governance pose similar challenges. The AI-driven and governed consumer services with interconnected healthcare systems will be positioned alike for novice consumers and expert healthcare professionals.
Ethical AI manifests these challenges, which often are expanded by social determinants of health. The nonmedical influencers embedded in AI for determining health outcomes are economic stability, education access and quality, healthcare access and quality, neighborhood and built environment, social, and community context. Governance and ethics go together for the purposes of monitoring AI services in production and assuring ethical AI services have control and respect and are regulated in a standardized matter. If a layer gets compromised or weakened by any means in deploying ethical AI services, it can have unintended consequences.
Potential risks of deploying AI services into healthcare management are risks that are related to AI: safety and security, bias, loss of autonomy, deception, deep fakes, discrimination, erosion of society, exclusion, incompetence, inequality, malicious use, violation or loss of privacy, loss of transparency, unintended consequences and even weaponizing of healthcare data. The behemoth of AI, which is making its way into healthcare services, needs consciousness and careful considerations of ethics, regulations, governance and the law. A global engagement from practitioners across multiple disciplines, such as AI ethicists, scientists, researchers, healthcare professionals and industry leaders will enhance the operationalization of ethical AI healthcare services.
The interconnected AI healthcare ecosystem will increase the complexity of healthcare systems consuming AI services if, at any time, the organization cannot explain the use and context of data. The trust, privacy, and safety and security aspects are closely interrelated with AI and ethics in the holistic delivery of AI healthcare services. Ethical AI frameworks and technology reference architectures will be designed and developed to include consumer controls, considerations, policies and guidelines.
In the Unite Paper | A Framework for Ethical AI at the United Nations, the framework encompasses the ethical AI to be trustworthy, explainable, transparent, responsible and beneficial. The consumer interest and benefits elevate once the implementations are governed with principles and consumer literacy is enhanced. This will gain the consumer trust in the interconnected systems and healthcare facilities.
The new generation of healthcare AI practitioners will manage and govern the ethical confidence and conscientiousness. Investments will pour in, consumer confidence will be elevated as ethical layers become transparent and it will provide scalability, growth and sustainability. Healthcare facilities will synergize with many other data producers who might be not only outside of the ethical boundaries and standards of healthcare operators such as the food and nutrition industry, suppliers and management companies, but will also start consuming services for messaging and delivering of healthcare events, and monitoring of security capabilities.
The data belongs to the patient, and they should have full control over use, release and monetization of their personal data. Similarly, algorithms leveraging AI inferences and blended in machine learning models need to contain ethical controls for effective governance. AI has the potential to positively impact the healthcare ecosystem and healthcare management if risk mitigation strategies are implemented and an adaptable ethical AI framework is implemented by multidisciplinary teams on a global scale.
Atul Gupta is Director, Enterprise Data Architecture, Sun Life, and former lead data architect, Merative
Utpal Mangla is general manager, Industry EDGE Cloud, IBM Cloud Platform.
Doreen Rosenstrauch, MD, PhD, FACHE, is founder, DrDoRo® Institute.