There are currently over 100 AI-driven drug discoveries in the pipelines of pharmaceutical companies worldwide. AI is also being used to create connected personal experiences for patients and providers, streamline regulatory reviews, personalise content, and develop hyper-informed treatment pricing models.

While classical AI has already revolutionised the R&D side of pharma, the focus is now shifting to the commercial side, where generative AI, or “right-brain AI,” comes into play. This type of AI is more intuitive and creative, capable of reasoning and understanding complex scenarios.

But, pharma is experiencing some of the scaling challenges that other enterprises face, as over 80% of companies are experimenting, but only 11% are achieving notable ROI, highlighting the nascent phase of this technology.

Twenty years later, we still aren’t anywhere near scaling autonomous cars on the road. Similarly, the evolution of e-commerce, with companies such as Amazon, which was founded in 1994, has seen only 25% to 30% of retail sales move online after 30 years. This doesn’t mean these technologies aren’t massively impactful.

As futurist Roy Amara observed, “We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run.” While the pharmaceutical industry is in the early innings of the AI wave, the long-term potential remains vast and transformative.

Responsibility involves establishing a solid foundation of governance and accountability. Every employee must understand the boundaries and goals of AI use within the organisation.

This means having a clear framework that outlines how the company will implement AI and what the boundaries are. This backbone of governance ensures that everyone is on the same page and that ethical considerations are at the forefront of AI implementation.

Reliability focuses on selecting use cases that not only are relevant, but also can clearly demonstrate the “why” behind them. It’s important to recognise that generative AI comes with a certain level of unpredictability. Hallucination, or the generation of incorrect or nonsensical information, is a feature, not a bug, of generative AI.

While we can curate data and build guardrails, there will always be some level of error. The key is to determine whether the organisation can tolerate this unpredictability in the chosen use case. Ensuring reliability means being prepared to manage and mitigate these risks effectively.

ROI should be considered early in the process, building at least a skeleton business case around all three factors: responsibility, reliability and ROI. This involves answering the “why,” giving people permission to explore and innovate, and providing them with the necessary governance framework. By doing so, the organisation can move forward, creating a culture that embraces AI while ensuring that it delivers real value.

The future of AI in pharma is promising, with the potential to free up human capacity for more value-added work such as creativity and strategy. Upskilling on both AI and soft skills, such as agility and critical thinking, will be crucial.

So, what’s the end goal? In my mind, it’s harnessing AI responsibly to drive real-world impact and improve patient outcomes. However, to do so requires careful implementation, collaboration and a focus on creating real value. By addressing the challenges and leveraging the opportunities, the pharmaceutical industry can transform its commercial model with AI, creating a better healthcare experience for everyone.