In this keynote session, Subrata Bose, PhD, Vice President, Diagnostic Imaging, Data & Artificial Intelligence, Head General Clinical Imaging Services at Bayer AG, discussed the recent innovations in AI in drug discovery, AI drug development, pharmaceutical AI, clinical trial AI and precision medicine AI. This session explored how artificial intelligence has transformed the landscape of drug discovery and development, moving beyond traditional trial-and-error methods to uncover novel drug candidates, optimize clinical trials and personalize treatments with unprecedented speed and accuracy.

Secure your place at The AI Summit New York to discover how industry leaders tackle pharmaceutical AI’s biggest challenges.

The 6 Key Takeaways from this Session:

Short on time? We’ve summarized the top 6 takeaways from the session below!

1. Traditional drug development’s $2.6 billion price tag and 90% failure rate create an innovation crisis that demands radical solutions - making AI integration not just valuable but essential for the future of medicine.
2. AI is transforming drug discovery from a costly gamble into a data-driven science, dramatically improving success rates at every stage from target identification to clinical trials. The technology won’t eliminate failure, but it’s making the path to life-saving medicines faster, cheaper, and more predictable than ever before.
3. AI transforms clinical trial recruitment from a costly bottleneck into a data-driven process, dramatically reducing timelines whilst improving patient safety and trial diversity.
4. AI’s pharmaceutical impact extends beyond discovery into manufacturing, supply chain, and regulatory processes. Success requires integrating these technologies across the entire drug development lifecycle, not just the laboratory phase.
5. Pharmaceutical companies face a triple challenge: integrating AI with legacy systems, proving ROI with traditional metrics, and managing novel risks like AI hallucination. Success demands new partnerships and evaluation frameworks built specifically for AI’s capabilities and limitations.
6. AI is compressing drug development timelines from decades to years, fundamentally altering how new medicines reach patients. The convergence of computational power, biological data, and regulatory evolution promises a new era of therapeutic innovation.

The below article has been created using a transcription of the video above, showing the session on AI in Drug Discovery and Development: Revolutionising Medicine which took place at The AI Summit London on June 11-12, 2025.

The $2.6 Billion Problem: Why AI in Drug Discovery Matters Now

Drug development has become an unsustainable gamble. Each new medicine requires $2.6 billion and 10-15 years of development, with only a 10% chance of success. The pharmaceutical industry screens millions of compounds to find just one that reaches patients – a process so inefficient it threatens medical innovation.

The numbers tell a brutal story. From 4 million initial compounds screened, only one typically receives final regulatory approval. Clinical trials alone devour 40% of total drug development costs. This attrition rate delays life-saving treatments from reaching desperate patients.

Subrata Bose, Vice President of Diagnostic Imaging, Data & Artificial Intelligence at Bayer AG, framed the challenge starkly at The AI Summit London. The traditional pharmaceutical development pipeline operates like a massive funnel where promising compounds disappear at each stage. What begins as thousands of potential breakthroughs narrows to hundreds, then dozens, until a single drug emerges – if lucky.

The human cost extends beyond balance sheets. Every failed compound represents years of research and billions in sunk costs that could have funded other innovations. Each delay means patients waiting for treatments that might never arrive. The industry needs a fundamental shift in how it discovers and develops new medicines.

Secure your place at The AI Summit New York to discover how industry leaders tackle pharmaceutical AI’s biggest challenges.

From Target to Treatment: How AI Accelerates Every Stage of Drug Development

Drug discovery has always been a numbers game - and the numbers have been brutal. Now artificial intelligence is rewriting the odds.

The pharmaceutical industry faces a stark reality: bringing a single drug to market takes over a decade and costs billions, with failure rates that would doom any other business model. Subrata Bose laid bare the challenge facing modern medicine at The AI Summit London. He explained how conventional approaches struggle with everything from identifying the right biological targets to predicting which compounds will actually work in humans.

The scale of inefficiency is staggering. Target identification - the crucial first step where researchers pinpoint which biological mechanism to attack - sees failure rates approaching 30%. Traditional screening methods fare worse, with hit rates hovering around 2%. By the time compounds reach lead optimization, half fail due to poor pharmacokinetics.

Bose detailed how AI transforms each stage of this gauntlet. Machine learning models sift through vast biological datasets to identify promising targets with unprecedented accuracy. Deep learning algorithms predict molecular behaviour, slashing the time needed to screen millions of compounds from years to weeks.

The impact extends beyond the laboratory. AI-driven approaches are fundamentally changing how pharmaceutical companies think about risk and investment. Companies can now fail faster and cheaper—a counterintuitive victory that accelerates the path to successful treatments.

Join The AI Summit New York and explore our conference schedule to learn more about transformative AI applications across industries.

Clinical Trial AI: Solving the Patient Recruitment Crisis

Clinical trials consume nearly half of all drug development costs, yet most fail at their most fundamental task: finding the right patients at the right time. This recruitment crisis has plagued pharmaceutical companies for decades, delaying life-saving treatments and inflating development costs. Artificial intelligence promises to transform how trials identify, engage and monitor participants.

Subrata Bose outlined the scale of the challenge at The AI Summit London. He described how AI systems analyse vast datasets of electronic health records, imaging data and genomic profiles to identify suitable trial participants with unprecedented precision (Bose, Vice President, Bayer AG).

The technology goes beyond simple matching algorithms. Modern AI platforms integrate multiple data streams to predict which patients are most likely to complete trials successfully. These systems analyse historical trial data to optimise protocol design, reducing dropout rates and improving statistical power.

Early implementations show promising results. Pharmaceutical companies report recruitment timelines cut by 30-50% when using AI-powered patient matching systems. The technology particularly excels at identifying candidates for rare disease trials, where traditional recruitment methods often fail entirely.

Experience cutting-edge AI innovations at The AI Summit New York by securing your registration today.

Beyond the Lab: AI in Manufacturing, Supply Chain, and Regulatory Approval

The pharmaceutical industry’s AI revolution extends far beyond drug discovery. While algorithms hunt for novel compounds in virtual spaces, the real test comes in scaling production, navigating global supply chains, and securing regulatory approval. These downstream challenges determine whether breakthrough medicines reach patients or remain laboratory curiosities.

Subrata Bose highlighted at The AI Summit London how pharmaceutical companies face mounting pressure to modernise their entire value chain (Bose, Vice President, Bayer AG). Manufacturing facilities designed decades ago must accommodate precision medicines requiring real-time quality control. Supply chains stretched across continents need predictive analytics to prevent disruptions.

The manufacturing floor presents unique AI opportunities. Computer vision systems detect microscopic defects in tablet production that human inspectors might miss. Machine learning models predict equipment failures before they disrupt production schedules.

Supply chain intelligence has evolved from reactive to predictive. The COVID-19 pandemic exposed vulnerabilities in just-in-time pharmaceutical logistics, forcing companies to build AI systems that anticipate disruptions. Modern platforms track everything from raw material availability to geopolitical risks, creating resilient networks that adapt to changing conditions.

Perhaps the most transformative shift occurs in regulatory submissions. The FDA and EMA increasingly employ AI to review the thousands of pages in drug approval dossiers. This creates a fascinating dynamic where AI-discovered drugs meet AI-powered regulatory review, potentially compressing approval timelines from years to months.

Ready to explore AI’s transformative potential? Book your spot at The AI Summit New York or review the comprehensive agenda featuring industry-leading speakers and sessions.

The Reality Check: Implementation Challenges and ROI for Pharmaceutical Companies

Pharmaceutical giants are discovering that AI’s promise comes with a hefty reality check. The gap between laboratory breakthroughs and boardroom returns is wider than many executives anticipated.

Subrata Bose laid bare the industry’s struggle with implementation at The AI Summit London. He highlighted how data integration remains a fundamental barrier (Bose, Vice President, Bayer AG). Legacy systems built over decades resist modernisation, while new AI tools demand standardised inputs that simply don’t exist.

The ROI question looms large. Traditional pharmaceutical metrics—time to market, clinical trial success rates, manufacturing efficiency—don’t capture AI’s diffuse benefits. Companies pour millions into AI initiatives without clear frameworks for measuring success.

Perhaps most concerning is AI’s tendency to hallucinate in pharmaceutical contexts. When algorithms generate plausible but incorrect molecular structures or suggest drug interactions based on spurious correlations, the consequences extend beyond failed experiments. Patient safety hangs in the balance.

The path forward requires rethinking partnerships. Academia’s role as an innovation engine becomes crucial when internal R&D hits these walls. Universities provide the experimental freedom that risk-averse pharmaceutical companies cannot afford.

Keen to hear more transformative AI discussions? Discover our comprehensive pass options, complete your registration, or preview the agenda for New York.

The Future of AI-Driven Therapeutics: From Bench to Bedside Faster

The pharmaceutical industry stands at a critical juncture. Traditional drug development takes 10-15 years and costs billions, yet artificial intelligence promises to compress these timelines dramatically. Industry leaders at The AI Summit London revealed how machine learning is already reshaping the journey from laboratory discovery to patient treatment.

Subrata Bose outlined the transformation underway in pharmaceutical research. The technology identifies promising compounds faster, predicts clinical trial outcomes more accurately, and streamlines regulatory submissions (Bose, Vice President, Bayer AG). What once required years of manual analysis now happens in months.

The impact extends beyond speed. AI-driven approaches are uncovering treatments for previously intractable diseases. Machine learning models analyse vast molecular databases, spotting patterns human researchers might miss.

Regulatory frameworks are evolving to match this pace of innovation. Health authorities recognise that AI-assisted drug development requires new evaluation criteria. This regulatory adaptation could prove as transformative as the technology itself, enabling faster patient access to breakthrough treatments.