Drug repurposing in the age of AI: lessons from COVID-19 and the future of second medical use innovation

In an environment where de novo drug development remains costly, time-consuming, and high-risk, drug repurposing in the age of AI offers a more efficient route to market by leveraging established safety and pharmacokinetic data.

What began as a largely serendipitous process, driven by unexpected clinical observations, has evolved into a data-driven and technology-enabled strategy. Increasing R&D costs (the average R&D cost to progress a drug from discovery to launch increased from an estimated $2.12 billion in 2023 to an estimated $2.23 billion in 2024 across 12 leading global biopharma companies), coupled with relatively low success rates for novel drug discovery, have pushed the pharmaceutical industry to extract greater value from known compounds and existing datasets.

At the same time, regulators, including the U.S. Food and Drug Administration, are becoming more open to repurposing, particularly where it addresses unmet clinical needs. Initiatives aimed at updating drug labelling and encouraging identification of new uses reflect this shift.

This transformation has been accelerated by two defining forces: the COVID-19 pandemic and the rapid maturation of artificial intelligence (AI). Together, they have fundamentally reshaped how new therapeutic uses are identified, validated, and brought to market.

The COVID-19 catalyst

The COVID-19 pandemic marked a turning point for drug repurposing. Faced with an urgent global health crisis, researchers and regulators turned to existing drugs as the fastest route to treatment.

Remdesivir, a drug initially developed or investigated for hepatitis, Ebola virus and Marburg virus, demonstrated how rapidly known compounds could be redirected to address emerging diseases. The pandemic triggered a global effort to screen existing drugs for antiviral activity, representing a shift away from traditional drug discovery models.

The global effort to screen existing drugs for new indications gave rise to a significant shift from traditional drug discovery approaches:

• Repurposing became large-scale and coordinated, rather than opportunistic
• Existing clinical data were rapidly mobilised
• Regulatory pathways were adapted to support faster deployment

Importantly, the global pandemic highlighted a core value of drug repurposing: that meaningful therapeutic advances can be achieved without developing entirely new chemical entities. However, it also exposed a key limitation. Many early repurposing candidates—particularly those identified through preliminary or purely computational approaches—failed to demonstrate efficacy in clinical trials. This underscored the need to balance speed with robust scientific validation.

The rise of AI-driven drug repurposing

If COVID-19 created urgency, AI has provided the tools to respond at scale.

AI-driven approaches use computational methods to identify new therapeutic uses for approved drugs by analysing large, complex datasets, including genomics, proteomics, clinical outcomes, and drug–target interactions. These approaches typically combine:

• Network-based modelling, which maps relationships between genes, proteins, and disease pathways
• Similarity-based analysis, which identifies drugs with comparable molecular or phenotypic profiles
• Machine learning models, which integrate multi-layered biological data

As a result, drug repurposing has shifted from reactive discovery to systematic hypothesis generation, significantly increasing the number of potential new indications. This evolution is enabling a more continuous model of innovation, where new uses for a drug can be explored throughout its lifecycle, rather than as isolated discoveries.

Drug repurposing in the age of AI: a new set of IP challenges

While AI is expanding the scope of drug repurposing, it is also reshaping the intellectual property landscape. One of the most immediate challenges arises from timing. AI systems can identify repurposing candidates at an early stage—often before supporting experimental data is available. This creates a familiar but increasingly acute dilemma: filing early to secure priority, with limited supporting data; or delaying filing to strengthen the application, at the risk of loss of novelty or third-party disclosure.

In parallel, the speed and scale of AI-driven discovery increase the risk of premature public disclosure, particularly where findings are generated in academic or collaborative settings. Early coordination between research and IP teams is therefore essential to preserve patentability.

More broadly, while the identification of repurposing candidates has become faster and more systematic, validation remains the critical bottleneck. Computational predictions alone are unlikely to satisfy patentability requirements, particularly where a credible therapeutic effect must be demonstrated.

Reconnecting innovation with protection

In this evolving landscape, intellectual property plays a central role in translating scientific insight into commercial reality. Unlike novel drug discovery, where protection is anchored in the compound itself, drug repurposing relies primarily on second medical use patents, which protect the new therapeutic application of a known substance. These patents are not ancillary—they are often the main source of exclusivity and value.

This reliance on use-based protection creates inherent vulnerability. The underlying compound may be off-patent or approaching expiry, limiting the ability to exclude competitors except in relation to the specific claimed indication. Accordingly, precise and enforceable claim drafting becomes critical.

How IP works in drug repurposing in the age of AI

Second medical use claims are designed to protect the relationship between a known drug and a newly identified clinical benefit. However, this relationship must be clearly and credibly established.

Patent offices are increasingly focused on whether applications demonstrate a specific and plausible therapeutic effect, particularly where the compound’s mechanism of action is already known. In the context of AI-driven discovery, this places greater emphasis on supporting computational predictions with a coherent scientific rationale and, where possible, experimental evidence.

Claim scope also requires careful calibration. Overly broad claims may be vulnerable to validity challenges, while narrow claims may be easy to design around. Given that AI may generate multiple potential indications for a single compound, applicants must make strategic decisions about which indications to prioritise, how broadly to claim them and how to build valuable and layered protection across a portfolio.

Conclusion

The future of drug repurposing will be shaped by the convergence of scientific capability, regulatory support, and intellectual property strategy. AI and data-driven approaches are rapidly expanding the range of potential therapeutic uses for known compounds. However, it is IP that determines which of these opportunities can be translated into meaningful patient and commercial outcomes.

For innovators, the key lesson is clear: in an environment where new indications can be identified faster than ever before, IP strategy must be integrated from the outset. The ability to secure, defend, and strategically deploy second medical use patents will be a defining competitive advantage in the next generation of life sciences innovation.

If you are exploring drug repurposing opportunities and want to ensure that your intellectual property strategy is robust, our team can help you identify and protect valuable second medical use claims, formulation innovations, and new dosing or administration routes. We can guide you through the complexities of securing patent protection while managing prior art risks and regulatory considerations. If you would like to discuss your project or have any questions about IP strategies for drug repurposing, please contact our specialists.

If you are concerned you may need some support to ensure that your intellectual property strategy is robust, our team can help you identify and protect valuable second medical use claims, formulation innovations, and new dosing or administration routes. We can guide you through the complexities of securing patent protection while managing prior art risks and regulatory considerations. If you would like to discuss your project or have any questions about IP strategies for drug repurposing, then please contact us today.

If you would like to find out more about this subject, the following reference points may be of interest:

• Intellectual property (IP) for life sciences
• FDA advances drug repurposing to address unmet medical needs
• Drug repurposing
• AI drug repurposing technology landscape 2026

If you are exploring drug repurposing in the age of AI or second medical use strategies, please contact Naomi Cheong or Fiona Law, in our Life Sciences team. Naomi, Fiona and their team can help you translate data-driven insights into robust, defensible IP.