Close this search box.

Is Africa ready for AI in pharmaceutical research and development?

AI developers on the continent should be supported to become competitive on the global stage.


Artificial Intelligence (AI) refers to the ability of algorithms encoded in technology to learn from data so that they can perform automated tasks without explicit programming of every step by a human. Given the recent explosion of interest in this new technology, it is clear that AI is here to stay, and it is being used in a wide range of applications, including in the research, development and delivery of pharmaceuticals. Although no new compounds that are fully identified or designed using AI have received approval for human use, it is important for Africa to be proactive about AI use in pharmaceutical research and development. Here is why.

A wide range of opportunities

AI has been found to be useful in most phases of pharmaceutical development. To put its fast-growing use into perspective, as of February 2022, there were 158 AI-based pharmaceutical discovery programmes and 73 AI-derived compounds in clinical development. What is more, there are 600 AI-based biotechnology and pharmaceutical companies, while US$115.84 billion has been invested in AI for pharmaceutical development between 2012 and 2022. AI is being used in basic pharmaceutical research, pharmaceutical discovery and design, such as in the development of therapeutic mRNA vaccines against cancer, preclinical studies to support or automate clinical development, patient recruitment for clinical trials, filing for regulatory approval and monitoring the safety of products on the market. Experts predict that in the (near) future a significant number of pharmaceutical products on the market will have been benefitted from AI during development, approval or marketing. For all these reasons, African countries must be proactive in terms of AI adoption, regulation and governance.

Second, the speed of drug development can be increased using AI because scientists will be able to predict how compounds under investigation will behave in the body and discontinue research on compounds that have a low probability of success. By failing fast, pharmaceutical companies will save on drug development costs, which can quickly add up to billions of US dollars which could be passed on to patients and be reflected in the price of medicines. While I am sceptical that this will be the case in the private sector, where profit is the main objective, state-funded research could provide an alternative to the private sector.

Another interesting use of AI is to find new uses for medicines that have already been approved. For example, baricitinib was identified in four days and subsequently recommended by the World Health Organization to treat patients with severe or critical COVID-19 after an AI-based company fed the information about how SARS-CoV-2 acts into its algorithm, which then searched at least 50 million articles in medical journals to identify the biological pathways that should be targeted to find a product that can be repurposed.

Furthermore, pharmaceutical companies are using AI for supply chain management, demand forecasting, shortage and stockout detection to design marketing strategies and decide prices for new medicines that will ensure maximum profit. To put it mildly, the possibilities are limitless.

Risks of using AI in pharmaceutical development and delivery

The lack of regulatory frameworks in Africa presents various risks. We know that pharmaceutical development has historically had ethical challenges and risks that will potentially be exacerbated by AI use. For example, it was reported in a study conducted in 2014 that White people made up 86% of clinical trial participants while another study conducted in 2019 found that 79% of genomic data was from people of European descent. AI technologies used in healthcare often use such data that exclude certain segments of the population, including African people, and will replicate these biases in their output.

Moreover, the inequities that exist in pharmaceutical R&D might be widened due to pharma and tech companies focusing on profit generation. For one thing, AI is expected to enable the pharmaceutical industry to develop personalised medicine (treatment customised for an individual patient). Although personalised medicine provides better outcomes for individual patients, they are available and accessible to only a few privileged people, which will increase health inequality. For another, pharmaceutical companies and researchers are already reported to be developing treatments for disease areas that have adequate data or are financially rewarding. This prevents the development of medicines that would address an unmet need (e.g. neglected tropical diseases) or be used by patients in low-income countries. African countries could prevent their own R&D programmes from taking a similar path.

Our countries must also address data use risks linked to the collection of consumer health data by tech and pharma companies. These risks range from the sale of data to third parties for commercial purposes, privacy violations because of difficulties in making data anonymous, to the potential for data leaks and cybersecurity breaches.

Other challenges relate to intellectual property governance and governance of the private sector. Algorithms used in drug development could be patented, which would prohibit other actors, such as academic researchers, NPOs and low-income country entities, from using them to improve drug development. The patents have the potential to prevent our R&D programmes from developing products for unmet medical needs that Big Pharma is not interested in. For Africa to reap the benefits of AI, these ethical challenges and risks must be addressed.

Proactive steps for African duty bearers

Beyond the needed introduction of standards, rules, regulations and legal frameworks to govern the use of AI in their jurisdictions, African governments must insist on thorough testing of algorithms in our settings, including testing of the AI technology’s assumptions, operating procedures, data properties and output decisions.

The need for human oversight in all steps of pharmaceutical R&D cannot be overemphasised. Additionally, AI developers on the continent should be supported to become competitive on the global stage. They are more likely to develop medical products that ensure equitable access in Africa and contribute to addressing the unmet needs of neglected populations.

National medicine regulatory authorities must also introduce or amend existing regulatory approaches and standards to ensure quality, safety and efficacy/performance of medicines developed using AI and AI-enabled medical devices. Collaboration across government departments will be critical to ensure the lawful collection of data and compliance with international data protection standards.

As the field evolves, it is important that countries on the continent do not implement different standards, as it will create a challenge for regulators who assess AI technologies and the developers using AI. In other words, Africa-wide initiatives should be pursued in the regard. For example, AU’s African Medicines Agency can step in to formulate new regulations to cater to the interests of Africans in the area of AI and medicines.

In sum, Africa has a choice to make. We can experience the development of AI in all sectors as a source of many risks or limitless opportunities for development.

Leave a Reply

Your email address will not be published. Required fields are marked *

Support The Pan African Review.

Your financial support ensures that the Pan-African Review initiative achieves sustainability and that its mission is shielded from manipulation. Most importantly, it allows us to bring high-quality content free of charge to those who may not be in a position to afford it.

You Might Also Like