The role of AI in research and development (R&D) is increasingly multifaceted. Beyond the technological possibilities, there are critical human, ethical, and sustainability considerations that must guide how we adopt and apply these tools. It’s no longer a question of whether AI is essential – it is. The real challenge is understanding how to use it responsibly.

Modern challenges require modern solutions

I once believed that scaling AI in R&D was simply a prudent strategy for biotechnology and pharmaceutical (biopharma) companies to remain competitive and innovative. Now, it’s clear that embracing this technology is essential. Organizations that fail to do so risk falling behind.

Focusing solely on efficiency gains from AI is insufficient. In an era marked by political polarization, economic uncertainty, and environmental degradation, we must adopt a holistic perspective on how AI influences (and is influenced by) the broader world.

Given AI’s environmental impact and its potential to disrupt the workforce, considering the human element is as important – if not more so – when deploying AI transformations at scale. In biopharma, this means prioritizing patient outcomes, employee wellbeing, and planetary health.

Biopharma companies cannot afford to sit out this technological revolution. But humankind cannot afford a slapdash, ill-conceived transition. Fortunately, there is a path forward for scaling AI responsibly.

A major challenge for R&D in biopharma

R&D is the engine that brings new drugs and therapies into existence. But steadily declining returns on investment for this vitally important stage have disincentivized proper financing levels.

Eroom’s law (“Moore’s” law in reverse) observes that new drug approvals have declined by 50 percent per inflation-adjusted spending every nine years since the 1950s. This trend was first formally described in a Nature Reviews Drug Discovery article from 2012 but has continued unabated.

At best, this makes it difficult to trial new, potentially life-changing treatments and, at worst, it prevents them from reaching the market altogether.

But recent advancements in AI and Gen AI have given biopharma companies the opportunity to identify and smooth out shortcomings and inefficiencies in the R&D process. This should incentivize proper levels of investment.

Bringing new drugs to market with AI-powered solutions

According to a new Capgemini Research Institute report, Smart bet, only option, or both? Biopharma R&D turns to AI , an overwhelming majority (82 percent) of biopharma executives think AI will fundamentally transform R&D for the industry. Furthermore, 63 percent agree that companies failing to scale AI will fall behind in innovation and market relevance.

The CRI report explores how, in the years ahead, various types of AI (e.g., machine learning, agentic AI) will transform every element of the R&D value chain: discovery, preclinical testing, clinical trials, clinical manufacturing, and regulatory approvals.

Research revealed that the most widely adopted AI use case in the drug discovery phase was target identification: finding the specific molecule involved in the development of a disease (such as a protein or gene) that a new drug will act upon.

Forty-three percent of organizations are already implementing AI in this way. Of these, 32 percent said that the process is more efficient and takes less time, reporting an average time savings of 28 percent compared to before adopting the technology.

What scaling responsibility with AI looks like in biopharma

The Pharmaceutical Research and Manufacturers of America (PhRMA), a trade association for research-based biopharmaceutical companies in the US, has committed to advancing innovation, making medicines more affordable, and creating a more just system.

AI could play a major role in achieving each of these goals – by optimizing complex R&D workflows.

Organizations can use AI to forecast the main results of clinical trials (i.e., whether a new drug will be effective or safe), create dosage schedules for patients, select the best locations for conducting trials, and predict who is most likely to experience side effects.

These companies can use AI to transform regulatory submission and approval workflows by automating the compilation of necessary data from internal and external sources, anticipating and answering regulator queries preemptively, and enhancing the overall quality of all documents submitted.

All these improvements don’t end with making pharmaceuticals more profitable. They get important medicines into the hands of those who need them most, without logistical hurdles or decreasing budgets getting in the way.

Where do we go from here?

Companies will need to make sure their technologies and cultures are prepared for AI adoption.

The previously mentioned CRI report shows that most organizations are underprepared in terms of data integration and AI training. Despite a majority having established data foundations, they still lag in building comprehensive data capabilities and are not yet operationally ready to scale AI.

AI-powered solutions can accelerate drug research in many ways but still need talented people who know how to use them properly. Human oversight will always be necessary for drug development. That’s not going to change.

It’s easy to discuss business transformation in abstract terms; it’s much harder to implement meaningful change that benefits both people and the planet. When approached thoughtfully, AI can achieve precisely that.