The IQVIA Institute for Human Data Science’s Global R&D Trends 2026 report found that by most measures, biopharmaceutical R&D was resilient in 2025. Funding stayed well above pre-pandemic levels. Seventy-nine novel active substances reached patients globally. Dealmaking remained active, with China-linked transactions reaching an all-time high.

End-to-end development timelines grew longer, matching the lengthiest point of the past decade and affirming that resilience and efficiency don’t always move in lockstep. Even so, industry-wide development success rates are holding steady — and we observed a tantalizing benefit for AI-enabled programs in one segment of the industry.

The Impact of AI Enablement

The Global R&D Trends 2026 report includes a comparison of clinical success rates for AI-enabled programs at emerging biopharma companies (EBPs) versus non-AI-enabled programs at companies of the same segment and size. For the most recent three-year window, the Phase I success rate for AI-enabled emerging biopharma programs was 75%. That is a substantial advantage over comparable non-AI-enabled programs, and a striking result for Phase I, where attrition has historically been much higher.

The Phase II picture also matters. If AI were simply accelerating programs into Phase I that would ultimately fail later, Phase II success rates for the AI-enabled cohort would drop, creating a poor outcome for overall R&D efficiency. Instead, the data shows that Phase II success rates for the AI-enabled cohort track are on par with their non-AI-enabled peers. Phase III transitions are not yet numerous enough to analyze meaningfully.

Two caveats are important. First, the validated cohort is relatively small, as the analysis draws on a carefully vetted set of programs with verified AI involvement rather than a broad self-reported sample. Second, while the signal is clear within emerging biopharma, industry-wide success rates are unchanged from the prior year. The improvement is visible in a segment, not yet in the whole.

Where EBPs Are Applying AI

Looking at the programs behind the success rate analysis, the largest share involves AI use for molecule discovery (that is, employing AI platforms to design a drug candidate, typically against an already known biological target). A smaller but more scientifically novel share involves target discovery, in which AI contributes to identification of the biological mechanism itself. A handful of programs combine both, using AI to identify the target and to design the molecule that acts on it. Indication selection — applying AI to determine which disease or which patient subset a molecule should be developed for — accounts for a smaller portion of the current cohort.

The 2025 deal data points in the same direction. Most of the AI-related R&D deals pharmaceutical companies signed in 2025 were discovery-related, predominantly for small molecules and, increasingly, for RNA-based therapeutics. The capital and the success signal are pointing in the same direction.

How Regulatory Signals Align

The regulatory side of the ecosystem is responding in step. The FDA is modernizing how it uses AI-based technologies and has released draft guidance on sponsor use of AI in drug development submissions. In addition, senior agency leadership has identified AI as a top priority in the New England Journal of Medicine.

Outside the U.S., Japan’s Pharmaceuticals and Medical Devices Agency has created a Chief AI Technology Officer role and published an AI action plan outlining how the agency will use AI to strengthen its own regulatory capabilities. Similar efforts are underway in European countries, too.

The parallel movement on the regulatory side reinforces what the success rate data is starting to suggest: AI’s role in drug development has moved from prospective to operational.

From Signal to System

In 2025, a combination of lengthening trial durations (with enrollment a critical contributor) and increasing inter-trial intervals have driven a notable slowing of overall development duration. There’s no single answer to the question of why investment and innovation momentum are not already translating into faster, more efficient development at an overall industry level.

For now, however, there’s evidence of AI’s impact — not on overall speed or efficiency but on success rates among EBPs. Not surprisingly, early adoption is showing up within these smaller, AI-native companies, where technology platforms and R&D strategy are built together from the start. How quickly that translates into industry-wide improvement in success rates depends on how effectively the broader ecosystem operationalizes what’s working.

Dive deeper into the findings and analysis in the full report, Global R&D Trends 2026: Advancing Innovation in a Changing Landscape.