Artificial Intelligence (AI) is gaining traction across a variety of sectors. The pharmaceutical industry is no exception to this. AI can be used for a variety of goals and is likely to become a staple in drug development, among other things. At the moment, many processes in the pharmaceutical industries take a long time and can require a high level of investment, so anything that can improve it and make it more efficient is welcome, but AI offers unique opportunities in this regard.

Drug development is the process of bringing a novel drug from “bench to medicine cabinets”. Discovering new drugs and bringing them to the market, as well as finding new purposes for existing medications, is a slow process and takes a lot of money as well. AI can improve the process in terms of speed and costs as well, making the clinical cycle shorter and more efficient. (1) Reducing time means that pharmaceutical companies and medical supplies suppliers can put a drug faster without losing quality or even improving quality, and the degree to which the drug development process can be difficult is staggering. Drug safety is the paramount concern, of course, which means that a product must be tested and tried before it can be used to recoup the investment. While the reasoning behind this is clear, there is still a lot of difficulty involved in putting out something new or repurposing an existing medication.

Regularly, it takes an average of 12 years, costs around $2 billion dollars for a successful drug to go from initial laboratory research to final sales in a medicine cabinet,  accounting for failures. The success rate is around 12%. AI can be 250 times more efficient than the traditional methods by increasing speed and accuracy, diversifying drug pipelines, and making the process safer by anticipating and considering potential side effects. (2) It is already possible to see this in action as more and more drug makers have employed AI methods to enhance the process of drug development and drug re-purposing.

As AI improves, COVID-19 has opened the door for various AI applications. First, it can be used for identifying new potential drugs for treating COVID-19, selecting FDA-approved drugs that can potentially help with the disease. But going beyond this, AI can be used for finding a new way to manufacture existing drugs, helping create new approaches that save money and ensure sufficient availability. (3) In addition, Randomized controlled trials have been used to verify AI-based findings for existing medication,(4) which suggests a new effective avenue for evaluating millions of drugs. Moreover，Covance (a US-based contract research organization and drug development services company) has tried to develop new automation solutions by streamlining processes and prototyping automation tools, leading to efficiency gains of about 15 to 25%, resulting in a million dollars in savings while improving quality. (5) As AI becomes more advanced, it is possible that the savings can increase. Something that distinguishes AI is machine learning that can improve through practice and become more sophisticated at the task at hand.  At the moment, there is a need to examine several hundreds of drugs to see if any might be used for COVID-19, and AI is uniquely positioned to help make this process faster.

Also, a National Institutes of Health (NIH) study proposes to examine more precise diagnostic biomarkers and drug targets for Alzheimer’s disease, using AI tools to examine the data. It will be the largest study (over 60,000 Alzheimer's patients) of its kind and an unprecedented attempt at uncovering new treatments for such a serious condition. (6) AI can make the process of data analysis faster and much more efficient, allowing researchers to identify biomarkers that could lead to potential treatments but could be difficult to detect in a traditional process.