Global pharma uses AI to speed drug timelines, but adoption remains uneven

Pharma leaders at BioAsia said AI is accelerating discovery and documentation, doubling pipelines at some firms, though its impact on clinical costs and timelines remains uneven

Global pharmaceutical (pharma) multinationals are increasingly deploying artificial intelligence (AI) to shorten drug development cycles and speed up product launches, reshaping both research pipelines and commercial execution, senior industry leaders said on the sidelines of the BioAsia Summit in Hyderabad.

 

From early-stage discovery to regulatory documentation and launch readiness, AI is emerging as a central lever to scale innovation without proportionately increasing costs or timelines. Executives cautioned, however, that the gains remain uneven across the value chain.

 

At Merck Sharp & Dohme (MSD), AI has been embedded across discovery and downstream functions, enabling the company to expand its pipeline while compressing time to market in parallel. The company plans to roll out more than 20 new drugs globally in the coming years.

 

According to Anton Groom, chief AI officer at MSD, AI-led discovery has doubled the company’s global drug development pipeline over the past two years by speeding up target identification and enabling earlier assessment of safety and efficacy.

 

Beyond research and development (R&D), MSD is deploying generative AI–driven “content accelerators” to overhaul regulatory, medical, and patient-facing documentation. These tools are cutting approval documentation timelines from weeks to days, allowing multiple products to move towards the market simultaneously while maintaining regulatory compliance, Groom said.

 

A similar strategy is unfolding at Takeda Pharmaceuticals, where AI is being embedded directly into R&D workflows rather than deployed as an add-on. Takeda’s R&D leadership has described AI-native laboratories and redesigned discovery processes as central to improving productivity and raising the probability of technical success across its portfolio.

 

Sanjay Patel, senior vice-president and global head of data, digital and technology innovation capability solutions and services at Takeda, said India is emerging as a strategic capability hub in this shift. Takeda’s Innovation Capability Center in Bengaluru is building enterprise-scale digital, data, and AI capabilities that support research, manufacturing, supply chains, and patient-facing functions globally, drawing on India’s deep digital talent pool and innovation ecosystem.

 

The company added that capabilities developed in India are designed to scale globally through shared platforms, reusable intelligence, and a connected operating model. This enables faster execution and greater enterprise resilience. In this framework, India functions as a strategic capability hub within Takeda’s global network, supporting the responsible adoption of AI and large-scale innovation aimed at improving decision-making and patient outcomes worldwide.

 

However, not all companies are seeing immediate reductions in costs or development timelines, particularly in clinical trials. Miltenyi Biotec offered a more cautious view during discussions on the sidelines of the conference. Founder Stefan Miltenyi said that while AI is being used widely in areas such as molecular design, imaging, and data analysis, its impact on products currently reaching the clinic has so far been limited.

 

“AI is really exciting, and we are using it in more and more areas,” Miltenyi said, adding that many development processes still rely heavily on traditional, paper-based workflows. While AI is already outperforming humans in tasks such as analysing large volumes of imaging data and identifying complex biological patterns, its broader effect on clinical development costs and timelines is only beginning to surface.

 

According to Miltenyi, the biggest gains are likely to emerge over time, particularly in clinical trial protocol design, statistical planning, data management, and regulatory report writing. Wider adoption, he said, will also depend on how quickly regulators in Europe, the US, and India adapt to AI-supported development models.