The growing use of synthetic data in modern pharma

More than ever, pharma firms are under increasing pressure to harness the power of AI, advanced analytics, and machine learning. But there is a quiet risk in pharma's AI rush: the misuse of synthetic data.

Drawing on my first-hand experience of the industry's push to offshore talent in the 1990s, I believe there are lessons for pharma executives grappling with today's data strategy. Back then, the efficiency case was real, as were the savings. But as firms leaned into a capability they didn't fully understand, the costs boomeranged years later in people and work that had to be hired and rebuilt. The lesson wasn't that offshoring was wrong. It was that efficiency pursued without regard for what is being hollowed out creates problems that surface long after the savings are booked.

This same principle applies to today’s unfolding AI-led transformation, particularly in the uptake of synthetic data. Yes, synthetic data holds great promise. Used well, it can extend evidence, simulate scenarios, simulate respondents, and accelerate analysis. It can also help overcome some of the challenges posed by traditional data collection, such as privacy and access. Used as a substitute for primary research, however, the reality is that it can create a false sense of knowledge and hallucinations.

The data problem

A model isn't neutral intelligence that improves every input equally. It reflects what it has been trained on and what it can reach. In pharma, where decisions are expensive to reverse, the inputs are everything. As AI makes analysis faster and more accessible, competitive advantage increasingly shifts upstream to the quality and uniqueness of data feeding the model. The models will become commoditised. The inputs will not.

This is where synthetic data becomes particularly interesting and potentially problematic. Healthcare AI spending nearly tripled in a single year, from an estimated $480 million in 2024 to $1.4 billion in 2025. Anecdotal signals suggest a meaningful share of that spend is shifting from proven primary research towards internal builds, synthetic data, and increasingly sophisticated models on the argument that AI can serve as a substitute for physician, patient, and payer research. There are some camps that believe synthetic data can completely replace valuable insights from primary research. That argument has no empirical support, yet. The issue isn't whether synthetic data is useful. It is what question we are asking it to answer.

Start with the question

Consider the questions that brand teams need answered to make a launch call, respond to a competitive entry, or defend a resource allocation: How is a specialist reasoning through a narrow indication right now? How does a biologic-experienced patient weigh a treatment switch? Where does a payer stand on a newly approved agent this quarter, not last year?

That information isn't published anywhere, and it isn't sitting in a claims feed waiting to be scraped. A general-purpose model can synthesise what's already known, but it can't manufacture behavioural evidence that doesn't yet exist. It can tell you what a market looked like last year, and not why prescribing has shifted over the last six months among the specialists who matter to your brand. That kind of evidence has to be gathered directly from the field, kept current, and generated independently from the public and manufacturer sponsored information that increasingly forms the common knowledge base available to company sponsored AI systems. And one part of this is structural, not a matter of degree: at launch, there's no historical data to synthesise. A more powerful model operating without evidence doesn't create evidence, it creates a more sophisticated inference.

That distinction matters. AI is extraordinarily good at finding, connecting, synthesising, and interrogating information. But it is not inherently an evidence-generation system. When the question is what physicians, patients, or payers actually think and do today, somebody still has to create the evidence.

Where synthetic data can help

There is a strong case for synthetic data. It can extend existing evidence, simulate scenarios, fill gaps and stress test hypotheses, and help teams explore questions before investing in more expensive research. What it cannot do on its own is establish new empirical evidence about how real physicians, patients, or payers are behaving today.

In this way, its role should be complementary, rather than substitutive. A synthetic respondent can produce a surprising answer. What it cannot establish is whether that answer reflects a real physician, patient, or payer in the market. In our research, the surprise is often the whole point: the physician whose reasoning breaks the pattern, the patient barrier existing evidence didn't predict, or the market shift that only becomes visible when new data is collected.

AI systems largely reason from what has already been observed, documented, or generated. If organisations increasingly replace primary evidence with synthetic outputs derived from existing evidence, the system risks becoming recursive: increasingly sophisticated analysis of an increasingly closed information set. Primary research does something fundamentally different: it introduces new observations into the system.

What synthetic data can’t tell us

This isn't a case against synthetic data. It's a case against confusing synthetic intel with real-world evidence.

Use synthetic data to accelerate the timely mechanical work, explore scenarios, and generate better hypotheses. Then, invest the time and money saved into collecting differentiated primary evidence and applying expert human judgement to what it means.

In this way, the competitive advantage comes from not using synthetic data, primary research, or human expertise alone. Rather, it will come from combining all three: differentiated evidence that others can't simply prompt their way into, technology capable of extracting more value from it, and experts who understand what that evidence means.

The models will keep improving. Synthetic data will become more convincing. That makes the quality of the underlying evidence not less, but even more important. The risk isn't that synthetic data becomes too powerful. Rather, it’s that we forget what it cannot tell us.

About the author

Dan Barton is CEO at Spherix Global Insights, where he leads the company's mission to deliver actionable market intelligence and strategic insights to pharmaceutical, biotechnology, and healthcare organisations. A seasoned healthcare and life sciences executive, Barton brings more than 30 years of industry experience and a remarkable track record of driving growth, innovation, and operational excellence in the pharmaceutical industry. Prior to joining Spherix, Dan served as CEO of Forian Inc., as CEO of Medical Outcomes Research Analytics (MORA), and earlier senior leadership positions at some of the industry's most respected organisations, including nearly 18 years at GlaxoSmithKline. Barton is passionate about the role data and market intelligence play in helping pharmaceutical companies better serve patients and navigate an increasingly complex healthcare landscape.