What Global Drug Pipelines Reveal About the Future of Medicine

Global drug pipelines had a total of 22,940 drugs active at the beginning of 2026. It's something that is easy to overlook, but shouldn't be. All of those molecules are wagers placed years ago before anyone is likely to hear the name of the drug.

As long as you read it correctly, a pipeline is not a list. It is the most "transparent" leading indicator of the direction that medicine is taking, which is usually 5 to 10 years before a medicine appears on a pharmacy shelf. And all that pipeline is shouting is a pretty interesting story right now about the diseases that are getting the attention, the working technologies, and the countries that are now the breeding grounds for pharmaceutical innovation.

By 2032, sales of prescription drugs are projected to be more than $2 trillion globally. The choices being made now in pipelines will be defining the terms for pipelines for the future. This is what the data actually means.

The Clinical Drug Development Process, Quickly

It's important to start the discussion by determining what is actually taking place at each stage of the process, before diving into what's popular. The clinical drug development process takes a molecule through a relatively standardized series of gates:

Preclinical studies (lab and animal studies) to determine basic safety and biological plausibility. The bulk of the pipeline is still in the preclinical phase – over 10,929 out of 22,940 active drugs worldwide as of early 2026.

Phase I — Safety testing in people (typically in small groups) as opposed to efficacy testing.

Phase II – proof of feasibility and dose determination in a larger but small number of patients.

Phase III — large, usually multinational trials to establish that the drug is effective and safe enough to be approved.

Regulatory review and approval: The process of submitting to regulatory authorities such as the FDA, EMA, or China's NMPA, which determine the acceptability of the evidence for market authorization.

Phase IV — post-market monitoring, at which point the performance of pipelines in the real world begin to influence the next generation of pipeline investments.

That is the textbook version! None of the drugs go through this process alone, however. It is fighting a battle with thousands of other programs in the same stages, and often in the same disease areas, for capital, for trial sites, for patients, for regulatory attention. That's why pipeline density is of equal importance to the progress of any single drug. A full Phase III field in one indication means that it's a sign of validation – the science is playing out – and of risk – the market is going to be very crowded. Another thin pipeline is either a missed opportunity or a yet unsolved issue.

Reading the development process at scale involves understanding the pipeline — not just one drug, but the entire system.

Real Global Pipeline for 2026

The headline numbers should be considered for a while. The list of 22,940 active drugs is down slightly from a year ago, but analysts say that's because the drugs were normalized with data rather than any decrease in industry activity. In the coming five years an extra 70-80 new active substances are expected to be released each year, representing 350 to 400 new medicines for the international market.

Oncology continues to be the biggest segment of pharmaceutical R&D, but this segment of the pipeline has begun to level off. Despite an increase in Phase II and Phase III starts, the overall number of trial starts in oncology actually fell 4% from 2024 to 2025. This is not contraction, it is consolidation! The field is closing in on the most promising late-stage assets, not opening up to a wide array of early-stage ones.

However, all is not quiet on that front: neurology, immunology, metabolic disease and rare diseases are all seeing their fair share of growth. An exception to this would be obesity. The total development assets of the global market of obesity medicines stood at over 193 by late 2025, with the market size being $66 billion in 2025, expected to grow to $92 billion by 2026.No longer a niche category. One of the fastest growing therapeutic categories in the entire industry.

AI Is Transforming the Pipeline's Content — But the True Test Is Yet to Come

It's easy to either underestimate the potential of AI in drug development or overestimate its impact. The truth, which is a combination of the above and where the data is now in 2026, is somewhere in the middle.

Early this year, over 173 AI-generated drug programs were in clinical development, a significant increase from about 24 at the end of 2023. Of these, 15-20 are predicted to get into key trials in 2026, without a doubt. The most eye-catching number: AI-found molecules have achieved 80-90% success in Phase I trials, while the industry average is approximately 52% over the past. That's a significant gap and it's one of the more obvious of the early indicators that something real is going on, rather than hype.

However, no drug found by AI has been approved as a full-blown medicine by the FDA. It is expected to be the first such approval in 2026 or 2027. The only real question that remains an open one, and one that will be answered or not, is whether AI delivers on outcomes in Phase III, where stakes are higher and science is harder. Upstream there is much clearer progress, in terms of identification of targets, prediction of drug-target interactions and/or significantly reduced preclinical timelines, with one widely cited study from 2023 conceding up to 25 to 50% time and cost savings this far.

There is regulatory interest. In January 2026, the FDA and EMA released what they called “Guiding Principles of Good AI Practice in Drug Development,” which was created as a joint effort.It was a joint effort between the FDA and EMA that resulted in the publication of “Guiding Principles of Good AI Practice in Drug Development” in January 2026, signaling this is no longer a technology being observed from the sidelines, but rather one that regulatory bodies are actively developing frameworks around.

For those with a strategic interest in pipelines, the takeaway is this: when AI-powered programs grow to a significant proportion of the total pipeline, it's not just a fascinating topic for conversation, it's useful knowledge for outpacing the competition — whichever pipeline it is.

Innovation is Redrawn - China

In the marketing community, the following change may not be given the credit it deserves: China has grown from a place where it was most known as a factory and manufacturing base to a true owner of the original, first-in-class pipeline assets.

It is hard to overestimate the magnitude of the change. During the first quarter of 2026 alone, Western pharmaceutical companies inked nearly $2 billion in licensing agreements with Chinese companies. During the entire year, Chinese assets will make up over two-thirds of the total value of Western pharma licensing deals (up from less than 5% five years ago). The therapeutic areas for these are oncology (especially newer classes such as antibody-drug conjugates and bispecific antibodies), immunology and a growing tide of cardiometabolic transactions as companies try to secure a slice of the obesity pie.

It's a pretty simple equation: Chinese biotech companies have some real good science early on but many lack the trial infrastructure, global sales channels and regulatory experience to commercialize it on their own on a massive scale. This means that near-term partnerships with multinational pharma companies are more likely than standalone attempts at global expansion for companies to have a shot at success.

In the real world of Western pharma and business development, what matters to them is that a larger percentage of the pipeline activity that counts — the sort that indicates potential future licensing opportunities or new competitive risks — is now occurring outside of the standard US/EU trial framework. Some teams are discovering that they are missing a crucial piece of the puzzle because they only had access to pipeline visibility data.

One Drug, Many Diseases: The Rise of Platform Medicine

This, perhaps the most structurally interesting change in today's pipelines, is that therapies are developing as platforms for multiple disease indications, rather than one-shot single disease treatments.

The best example is related to obesity. The original diabetes treatments are now in the spotlight for their potential to treat cardiometabolic diseases and even psychiatric disorders, and are now becoming “whole-body” drugs rather than “weight-loss” drugs. These are the next generation of multi-mechanism, multi-indication assets from Novumodisk and Eli Lilly, respectively, which can be expected to enter the US market in April 2026.

This trend is also beginning to emerge in cell and gene therapy, which has been a very oncology-centric area of research. The data reflects actual movement: Where 39% of newly initiated gene therapy trials were indicated for non-oncology conditions in the previous year, that figure has risen to 51%.The data represents real movement – only 39% of newly initiated gene therapy trials were indicated for non-oncology conditions the year before, but that has now jumped to 51% – a significant structural shift in less than a year. In 2025, the first personalized in-vivo CRISPR base-editing therapy was given to an infant with a rare metabolic disease, known as "Baby KJ.Personalized in-vivo CRISPR base-editing therapy took a giant leap forward when it was given to the first infant with a rare metabolic disease in 2025, dubbed "Baby KJ. It also led to a new FDA development, its "Plausible Mechanism Pathway," announced in November 2025, specifically to speed up the approval process for ultra-rare and single patient therapeutics where traditional randomized trials simply are not possible.

The challenge that this poses in practice–one asset can hold a significant value in more than one therapeutic pipeline simultaneously, so previous indication-by-indication pipeline analysis is no longer complete. It is becoming more critical to have an understanding of a molecule's actual strategic value or the competitive crowding it will likely encounter, more than just the lead, initial indication.

What This Means in Practice

The data from the pipeline suggests that innovative pharma and biotech companies should focus on areas of true scientific traction (obesity, immunology and rare disease) over oncology's late-stage competitive field, as well as not discounting the deals acquired from China, and keeping an eye on the early-stage programs coming out of the AI world. Pipeline replenishment is not just a cyclical challenge, as more than $500 billion in branded drug sales will be at risk due to patent expiry on the horizon through 2030.

The patent cliff is an opportunity, not a threat, for generic drug companies. Finding out which molecules are coming into the arena of LOE, and what therapeutic areas are being opened up by molecules that are approaching LOE, is a direct pipeline intelligence exercise — obesity, for instance, with a number of molecules with patent expiration dates in 2026, is a live example right now.

The pipeline points clearly for CROs: the number of specialized, in most cases adaptive trial designs for rare disease and cell/gene therapy programs are increasing; large international trials are increasingly moving into Western pathways for obesity and cardiometabolic trials; and the need for cross-border regulatory expertise continues to rise as more China-originated assets begin to enter Western trial pathways after receiving licenses.

The Pattern Is Only Visible at the Pipeline Level

The path to developing any one drug does not tell you that medicine is going from treatment by disease to treatment by platform. No one has told you that, in the early stages, AI is quietly beating historical benchmarks, or that China is no longer just a place for manufacturing, but one of actual first-class science. The only way these changes are apparent is by looking at the pipeline as a whole, thousands of programs at each step, all at once.

That's why, today, structured, up-to-date pipeline intelligence has ceased to be a research convenience and has become a strategic necessity. But it isn't just one trial outcome that dictates the future of medicine. It's there on the pattern, just there, in thousands of them — and for the teams that have the data to see it.

Optimize Your trial insights with Clival Database.

Are you exhausted from the uncertainty of trial insights pricing? Clival Database ensures the clarity in the midst of the global scenario for clinical trials to you.

Clival Database is one of the best databases that offers an outstanding number of clinical trial data in terms of 50,000+ molecules and from primary regulatory markets as well as new entrants like Indian and Chinese markets.

Elevate your trial success rate with the cutting-edge insights from Clival database.

Check it out today and make more informed sourcing decisions! Learn More!