The field of Artificial Intelligence (AI) holds significant prospects in bolstering the efficacy of healthcare and medicine and as a result, a correlating potential of achieving the universal health coverage the world needs. In the field of Clinical Genomics for instance, which is the study of clinical outcomes with genomic data, AI techniques have emerged as valuable tools that effectively address the obstacles encountered by clinicians and genetic counsellors in dealing with the vast volume of genomic information and extensive testing reports inherent in comprehensive genomic sequencing. The integration of AI has significantly facilitated the process of data mining and integrative analysis, streamlining these procedures to a considerable extent. Examples of these advancements include refined diagnostic techniques, more effective disease treatment and management strategies, and accelerated drug discovery and development processes. Nevertheless, the integration of AI within healthcare instigates a range of ethical and legal dilemmas that necessitate thorough consideration. Specifically, the utilisation of AI in the field of clinical genomics, engenders various legal quandaries, encompassing matters pertaining to data privacy and security, protection of intellectual property rights, determination of liability in instances of errors or malfunctions, and adherence to regulatory frameworks.
Artificial Intelligence in Healthcare
The term “Artificial Intelligence” (AI) generally refers to the ability of computer programs to undertake tasks that are commonly associated with intelligent beings. At the core of AI lies algorithms —sets of instructions which are translated into computer codes that provide information for the rapid analysis and transformation of data, leading to the derivation of conclusions, information, or other outputs. An AI system refers to a machine-based system capable of making predictions, recommendations, or decisions that have an impact on real or virtual environments, all within the context of predetermined human-defined objectives. These AI systems are designed to operate with varying degrees of autonomy. AI technology encompasses diverse types, including machine-learning applications that involve tasks such as pattern recognition, natural language processing, signal processing, and expert systems. Machine learning as a subset of AI techniques, relies on statistical and mathematical modelling approaches to define and analyse data. Through this process, learned patterns are then applied to execute or guide specific tasks and generate predictions.
The contributions of AI to the healthcare industry are multifaceted and extensive. Among the manifold benefits, several noteworthy outcomes can be highlighted. Firstly, AI has played a pivotal role in empowering patients and communities, enabling them to assume a more active role in managing their own healthcare and enhancing their comprehension of evolving healthcare needs. Secondly, AI has significantly augmented the capabilities of healthcare providers, facilitating improvements in patient care through enhanced diagnostic accuracy, optimised treatment planning, and support for pandemic preparedness and response efforts. Additionally, AI has proven instrumental in informing the decision-making processes of health policymakers and assisting in the allocation of resources within health systems. Notably, AI has also emerged as a valuable solution for resource-constrained regions, particularly in low-resource countries where patients often face limited access to healthcare professionals. In such contexts, AI has bridged the gaps in healthcare access, enabling the equitable provision of essential health services. Notwithstanding the above benefits of AI, it is safe for one to conclude that indeed AI’s opportunities and challenges are inextricably linked.
How laws and policies impact the use of AI in clinical genomics and healthcare
AI technologies hold great promise for the healthcare industry and have already contributed to great advances in the field of genomics. However, in order to harness its full potential and ensure patients and communities meaningfully benefit from it, it is imperative to establish and enforce ethically sound laws and policies. The formulation and implementation of these policies will ensure the safety, accuracy and efficiency of AI systems and also promote adherence to instructional usage.
Engaging this line of action will also significantly build trust in AI technologies, guarding against negative or erosive effects, as well as the proliferation of contradictory guidelines. This responsibility not only rests on government agents and organisations but also on the designers and developers of such technologies, compelling them to meet specified standards that will enhance the overall governance and ethical integrity of AI deployment in the healthcare industry.
Over the years, efforts have been made to safeguard human rights through explicit legal mechanisms, evident in various international and regional human rights conventions we have today. These include prominent instruments like the Universal Declaration on Human Rights (UDHR), the International Covenant on Economic, Social and Cultural Rights (ICESCR — which encompasses General Comment No. 14 —the right to health), and the International Covenant on Civil and Political Rights (ICCPR).
Despite the robust human rights-oriented legislative frameworks, patients’ rights are still at risk of being undermined or redefined by global health corporations and governments.
Regional human rights conventions, such as the African Charter on Human and People’s Rights (ACHPR), the American Convention on Human Rights (ACHR), and the European Convention on Human Rights (ECHR), further contribute to the framework of human rights protection. Notwithstanding the robust human rights-oriented legislative frameworks, patients’ rights are still at risk of being undermined or redefined by global health corporations and governments. Putting into consideration the fact that AI performance largely depends on data collation, these entities have made it a culture to gather data and ultimately turn it into profitable commercial ventures.
Data Protection Framework
Data collection serves as the foundation upon which the field of Clinical Genomics relies, instrumental in the unravelling of intricate functional information embedded within DNA sequences. This in turn facilitates early disease diagnosis and the assessment of an individual’s predisposition to a particular disease. Engaging in genomic data science which involves the acquisition of vast datasets from numerous individuals, entails a range of ethical obligations. Unfortunately, it is evident that these responsibilities have been gravely disregarded and exploited in many instances. This, therefore, raises concerns that can only be met by a comprehensive legal framework that guides and regulates the manner in which government agencies and global health corporations collect and process individual data.
For instance, very recently in Nigeria, the Federal Government signed into law the Data Protection Act 2023, which has, as one of its primary objectives, the need to safeguard the fundamental rights of data subjects, including those from whom genetic information is obtained. Section 24(1)(f) of the Data Protection Act further imposes a responsibility on data controllers and data processors to ensure that personal data is processed in a manner that guarantees the appropriate security of those data, including the duty to ensure that they are adequately protected against unauthorised or unlawful processing or any form of data breach.
Again, Section 25(2) of the Data Protection Act upholds the principle of fundamental rights by limiting the interests of data processors and data controllers in personal data processing, if the same poses a threat to the fundamental human rights of the data subject. Another noteworthy provision within the Act is the establishment of the Nigerian Data Protection Commission, which assumes a pivotal role in overseeing various aspects of data management, including those within the healthcare industry. There is no doubt that this proactive step taken by the government would help in ensuring the responsible and judicious collection and utilisation of data by stakeholders in the healthcare industry.
Addressing the ethical, legal, commercial and social concerns
It is worthy of note that legislation focused on data protection alone would in itself be limited in effectively addressing the issues resulting from the use of AI in clinical genomics and healthcare. This is because the use of AI in clinical genomics and the healthcare industry, in general, raises transnational ethical, legal, commercial, and social concerns. To address these concerns, laws and policies would have to be strategically structured to ensure inclusiveness and equity in the use of AI.
This would entail that AI technologies are designed to encourage the widest possible appropriate and equitable use, irrespective of age, sex, gender, income, race, ethnicity and other characteristics protected under globally recognized human rights laws. Such strategic provisions would be directly aimed at ensuring equal access to Artificial Intelligence (AI) technologies, preventing the encoding of biases that perpetuate disadvantages for identifiable groups, particularly those already marginalised. In the context of clinical genomics and healthcare, it is imperative for AI technologies to actively mitigate and minimise inequitable power differentials that may arise between providers, policy-makers and individuals, as well as companies and governments involved in the development and deployment of AI technologies, and those who depend on and utilise them.
Conclusion
To prevent oversight and provide assurance of safety and efficacy in the use of AI in the field of clinical genomics and healthcare in general, government regulators should consider enacting laws that require the disclosure of specific aspects of AI technologies, while taking into account proprietary rights. This transparency may encompass elements such as the source code, data inputs, and analytical approach employed by the AI system. By requiring transparency, regulators can improve their ability to assess and monitor AI technologies.
In addition, government regulators should also establish requirements for rigorous testing of AI systems. When this is in place, highly reliable evidence can be acquired through randomised test trials. Relying solely on comparisons with existing datasets in laboratory settings may be insufficient. Robust testing methodologies, such as randomised trials, can provide much more reliable evidence on the performance and effectiveness of AI systems.
Furthermore, government regulators should incentivize developers to proactively identify, monitor, and address safety and human rights concerns during the design and development of AI products. By promoting the incorporation of ethical considerations from the early stages of development, regulators can encourage the responsible and ethical implementation of AI technologies.
Finally, government regulators should mandate or conduct robust market surveillance to identify biases in AI systems. This surveillance aims to detect and address potential biases that may arise from the use of AI technologies in healthcare. By actively monitoring the marketing and deployment of AI systems, regulators can safeguard against discriminatory outcomes and ensure equitable access and treatment.
Amala Umeike is a Partner at Stren & Blan Partners and supervises the Firm’s Health and Pharmaceutical Sector. Clara Eze is an Associate in the Firm’s Dispute Resolution and Intellectual Property Departments, while Emmanuel Ughanze is an Associate in the Firm’s Health and Pharmaceutical Sector.
FIRM’S BIO
Stren & Blan Partners is a full-service commercial Law Firm that provides legal services to diverse local and international Clientele. The Business Counsel is a weekly column by Stren & Blan Partners dedicated to providing thought leadership insight on business and legal matters.
Connect with Stren & Blan Partners:
Website: www.strenandblan.com
LinkedIn: linkedin.com/company/strenandblan
Twitter: twitter.com/Strenandblan
Instagram: instagram.com/strenandblan
