Month: March 2026

By Laura Farmer, President, Opus Strategy

JPM Week in San Francisco is always a useful barometer for the life sciences industry. Not for what’s “trending,” but for what is moving from possibility into execution. The conference itself matters, but the real signal often comes from the adjacent meetings, the smaller sessions, and the conversations that happen when people stop pitching and start problem-solving.

This year, the tone felt more constructive than it has in a while. Not exuberant. But steadier. Investors were asking more disciplined questions about differentiation, durability, and proof. Operators were sharper about what they need in order to win the next phase. And in many rooms, the conversation shifted away from “big ideas” toward pragmatic plans: how to generate evidence efficiently, how to scale manufacturing, how to secure coverage, and how to make adoption predictable.

For pharma leaders, that shift matters. It suggests the industry is entering a phase where the winners will be the teams that can translate scientific promise into repeatable value creation, across portfolios and across therapeutic areas.

As went to events and spoke with colleagues, there were three themes I heard repeatedly:

First, the funding environment is reopening, selectively. Capital remains scarce for programs with weak data or unclear clinical utility. But for companies with credible differentiation and a coherent path to adoption, the window feels less constrained than it did even 12 months ago. Importantly, “strong data” is no longer interpreted narrowly as a positive trial readout. It increasingly includes clarity on the patient definition, the biomarker strategy, the regulatory plan, and the pricing and access narrative. In other words, teams are being rewarded for reducing uncertainty across the full development and commercialization arc.

Second, large pharma is in pipeline-building mode, unmistakably. Between patent expirations and growth expectations, the pressure to build and buy is real. What stood out this year is how central precision medicine and novel modalities have become a pipeline strategy. These are not side bets. They are core to how many organizations are thinking about the next decade of growth, particularly in areas where payer and provider expectations are tightening and where “me too” products face a rapidly shrinking runway.

Third, scalability has become a litmus test. The industry is no longer satisfied with brilliant therapies that cannot be produced at scale, delivered efficiently, or reimbursed sustainably. The conversations are increasingly integrating manufacturing, logistics, evidence requirements, and workflow adoption into the scientific story. Breakthrough biology is necessary, but it is not sufficient on its own. If you cannot make the therapy, deliver it, and get it used in real-world settings, the science will not matter as much as it should.

Precision medicine as the connective tissue

These themes converge most clearly in precision medicine. Leading into JPM, the field has been moving from “promise” to more visible proof. We are seeing better multi-omics integration, improved biomarker rigor, and a faster feedback loop between discovery, development, and clinical decision-making.

AI is clearly a catalyst here, but the most credible narratives were not “we used AI.” They were “we used AI to reduce uncertainty.” Better patient selection. Faster identification of responders. Earlier signals. A more precise understanding of heterogeneity. Pharma teams have heard the AI hype. What they are looking for now is a measurable advantage that shows up in development timelines, trial efficiency, label strength, or commercial adoption.

At the same time, expectations for commercialization are rising. Precision medicine programs are increasingly expected to show early thinking on access, evidence, and adoption. The best teams are building with payer logic and clinical workflow reality in mind from the beginning, not bolting it on after publication. For medical directors and product teams, this is a meaningful shift. It changes when and how you engage on evidence generation, and it raises the importance of aligning clinical development with the future label story and the future value story.

A session that went deeper: Penn Center for Innovation on Precision Health

One of the most substantive discussions I attended was a Precision Health session hosted by the Penn Center for Innovation. It was a welcome contrast to the usual surface-level conversations and got into the hard truths that will shape what happens next. Three themes stuck with me:

“Best of times, worst of times” for academic medicine

Dr. Jonathan Epstein, Dean of the Perelman School of Medicine and Executive Vice President of the University of Pennsylvania for the Health System, was candid about the strain on academic research. Funding pressure is real, and it directly affects the biomedical pipeline that industry ultimately depends on. At the same time, the opportunity is enormous, especially as AI-enabled discovery and translation help compress timelines from insight to intervention. The tension is clear: the science is accelerating, but the ecosystem that generates and validates that science is under increasing stress.

For pharma leaders, this is more than a macro observation. It will shape access to partnerships, the pace of external innovation, and the availability of high-quality translational research. It also underscores the value of academic-industry collaboration models that are built for speed and rigor, not for optics.

Rare and ultra-rare therapeutics need a better commercial path

Dr. Peter Marks, Senior Vice President of Molecule Discovery and Head of Infectious Disease at Eli Lilly and former Director of the FDA’s Center for Biologics Evaluation and Research, highlighted a reality that the industry continues to struggle with: if we want a full complement of rare disease therapies, the path must become faster and less expensive without sacrificing rigor.

One idea that resonated was the concept of process-based approvals. Instead of forcing each molecular variant through a full regulatory “toll booth,” a validated end-to-end lab and manufacturing process could potentially be reused across individualized therapies. For regulated science, this is a meaningful reframing. It shifts the value from a single asset to a scalable platform and challenges us to rethink what we define as “the product” when therapies are increasingly personalized.

For medical and commercial teams, rare disease strategy has always required careful choreography between evidence, access, and trust. What changes now is the scale of the ambition. If the industry wants broader rare disease coverage, the system must evolve to support it. Otherwise, the economics will remain structurally misaligned with the clinical need.

Immune health as a transformational frontier, with AI as an enabler

Dr. Allison Rae Greenplate, Adjunct Assistant Professor of Systems Pharmacology and Translational Therapeutics and Director of Immune Health at Penn Medicine, described a vision that felt both ambitious and increasingly plausible: decoding an individual’s immune profile at scale and using it to match patients to therapies with far less trial-and-error.

This is where precision medicine starts to look less like a niche approach and more like a broader operating system for care. Stratify better. Treat smarter. Reduce waste. Improve outcomes that matter clinically and economically. For pharma, this is a reminder that the future of differentiation may be less about “a new mechanism” and more about “a better match,” supported by evidence and operationalized through diagnostics, data, and workflow integration.

The practical question precision medicine keeps forcing us to answer: How do we build trust?

Precision medicine is fundamentally a trust business. Trust that the data are sound. Trust that stratification is meaningful. Trust that outcomes translate to real-world settings, not just curated cohorts. This is where medical affairs and evidence strategy become central to success, not supportive functions downstream.

Why this matters now

The next phase of precision medicine will not be won by the loudest storytellers. It will be won by teams that translate complexity into clarity:

• Clear clinical utility
• Clear evidence strategy
• Clear market access pathway
• Clear operational plan for scale

The science of precision medicine is moving fast. The healthcare system is still the healthcare system. Translation is the work.

By Laura Farmer, Opus Strategy

Recently, I attended the AI in Medical Affairs Conference in Philadelphia, and came away with a clear impression: the conversation around AI in #MedicalAffairs has entered a more serious phase. The focus is no longer on potential or experimentation alone. It’s on execution: how organizations integrate new capabilities responsibly, sustainably, and in ways that truly support scientific and business objectives.

Here are the five lessons that stood out most, and that I believe matter most for leaders navigating this transition.

1. The challenge is no longer technological—it’s organizational

Most Medical Affairs organizations are past the point of asking whether AI belongs in their function. Tools have been piloted, use cases explored, and early value demonstrated. Yet many teams remain stuck in pockets of experimentation rather than moving toward scaled impact.

What’s holding them back is rarely the technology itself. Instead, the real work begins with aligning workflows and understanding decision rights, governance, and expectations. AI introduces new ways of working, but organizations often try to layer those capabilities onto existing structures and workflows without rethinking how work actually gets done. Leaders who recognize this as an organizational #transformation (not a digital add-on) are the ones seeing progress.

2. Quality of scientific exchange is becoming a deliberate capability

Several discussions focused on the evolving understanding of Medical Affairs interactions, particularly between MSLs and HCPs. These conversations were framed not as isolated events, but as cumulative experiences shaped by listening, empathy, credibility, and relevance over time.

Here’s what’s shifting: The ability to support this type of development around the interactions between MSLs and HCPs systematically. AI-enabled tools are beginning to provide structured feedback and insights that help teams reflect on how they engage, identify patterns, and improve consistency. When used thoughtfully, this kind of support elevates practice across a team without reducing interactions to scripts or metrics. It enables professional growth while preserving scientific judgment.

3. Adoption depends on leadership clarity and change discipline

Adoption emerged repeatedly as the differentiator between organizations making progress and those stalled in repetitive pilot modes or adoption purgatory. Tools alone do not change behavior. What does change behavior is clarity and connections with end users, understanding and educating teams on why a tool exists, how it fits into daily work, and what success looks like.

Organizations seeing momentum have invested in change management fundamentals: training grounded in real workflows, internal champions who model new behaviors, and leadership messaging that signals both permission and expectation to adopt. Without this foundation, even well-designed tools struggle to move beyond early adopters.

4. Governance is enabling trust, not slowing progress

A consistent theme across sessions was the importance of verification. AI-generated outputs are widely treated as inputs that require review, context, and traceability, particularly in regulated environments.

Rather than creating friction, clear governance and guardrails have helped teams move faster. When guardrails are explicit, users know where automation is appropriate and where human oversight remains essential. This clarity builds confidence, supports compliance, and reinforces scientific integrity. Governance, when designed intentionally, becomes an enabler rather than a constraint.

5. Value is being measured in capability, not just efficiency

Conversations around ROI reflected a more mature understanding of value. While time savings still matter, leaders increasingly emphasized outcomes such as improved quality, greater consistency across teams, and the ability to scale best practices.

These measures resonate because they connect #AI investments to long-term capability building. Leaders are asking whether AI helps teams make better decisions, engage more effectively, and operate with greater confidence—not just whether it speeds up existing tasks. This shift in measurement is helping organizations build more durable business cases and align investments with strategic goals.

What does this mean for Medical Affairs?

The organizations that will stand out in the next phase of Medical Affairs transformation are not those with the most advanced tools, but those that approach adoption deliberately. Success depends on leadership alignment, thoughtful change management, clear governance and communications, and metrics that reflect real value.

AI is becoming part of the operating fabric of Medical Affairs. The opportunity now is to shape how it’s embedded—intentionally, responsibly, and in service of both science and strategy.

What Leaders Should Do Next

For leaders, the next phase of AI adoption success by the Medical Affairs team will be more about intentional design than continual experimentation. A few actions can help organizations move from pilots to sustained impact.

1. Anchor AI initiatives in clear business and scientific priorities. Teams should be explicit about the problems they are trying to solve—whether improving consistency of scientific exchange, accelerating insight synthesis, or strengthening decision-making. Clarity at the outset prevents tools from becoming disconnected from real needs.
2. Invest in change management as deliberately as in technology. Adoption requires training that reflects actual workflows, visible leadership support, and internal champions who help translate strategy into practice. Without this foundation, even strong tools will struggle to gain traction.
3. Define governance early and communicate it clearly. Establishing expectations around verification, oversight, and acceptable use builds confidence and trust across teams. When guardrails are understood, users can engage more freely and responsibly.
4. Rethink success metrics. Move beyond short-term efficiency gains and track indicators that reflect capability building—quality, consistency, scalability, and decision support. These measures better capture the long-term value of AI in Medical Affairs.
5. Plan for how work will shift. AI changes not just how quickly work is done, but where time and expertise are required. Leaders who anticipate this shift and adjust capacity accordingly will be better positioned to realize meaningful returns.

The organizations that act with intention now—aligning leadership, people, and processes—will be the ones that turn AI into a durable advantage in Medical Affairs.

How change management determines whether AI becomes an enterprise capability or remains a permanent pilot.

By Laura Farmer, President, Opus Strategy

The biopharma industry has never had a shortage of ambition when it comes to AI. What it has a shortage of is follow-through.

According to a December 2025 Deloitte report, only 22 percent of life sciences leaders report successfully scaling AI initiatives beyond pilots, and just 9 percent report achieving significant financial returns.

The problem isn’t the technology. The tools work. The problem is organizational, and it’s one the industry continues to underestimate.

AI Transformation Is Different From Every Other Change Biopharma Has Managed

Most biopharma change is vertical. AI is horizontal.

Biopharma organizations are experienced with change. Asset launches, system implementations, and functional reorganizations are difficult, but they are contained. They occur within a function, a program, or a geography. The underlying operating model remains largely intact.

AI doesn’t work that way.

It cuts horizontally across the enterprise, touching manufacturing, quality, regulatory, clinical, and commercial functions simultaneously. It doesn’t simply introduce new tools; it changes how evidence is interpreted, how risk is evaluated, and who is accountable for decisions supported by algorithmic systems. That’s a fundamentally different kind of change, and the playbooks organizations have relied on historically don’t fully address it.

There is also a cultural mismatch at the center of this challenge. Biopharma organizations are built around stability. Protocols are locked. Processes are designed for audit readiness. Change control exists for good reason.

AI systems operate differently. Models update as new data becomes available. Managing that tension requires structured lifecycle governance for AI models, including validation checkpoints, documentation standards, and defined escalation paths. Not because bureaucracy demands it, but because without those guardrails, teams simply will not trust the outputs even when the underlying technology performs as intended.

Most Organizations Can Prove AI Works. Far Fewer Can Make It Stick.

Most companies can demonstrate value in an AI pilot. The problem is what happens next.

Pilots prove that technology works. They rarely prove that the organization is prepared to use it at scale.

Three structural breakdowns tend to appear repeatedly when organizations attempt to move from experimentation to enterprise capability.

The first is messaging that focuses on features rather than workflow. Teams learn what a tool does, but no one shows them what work will change, which tasks can be discontinued, or how daily decision-making will be affected. Without that clarity, adoption remains superficial. People use the new system when required and revert to familiar workflows when oversight disappears.

The second structural breakdown is late governance. Regulatory, quality, and compliance teams are often engaged only after pilots demonstrate value. This positions them as gatekeepers rather than collaborators. Governance introduced after the fact creates friction. Governance embedded from the beginning creates confidence.

The third, and least discussed, is the accountability gap. Biopharma professionals remain personally responsible for the rigor and defensibility of their work. When AI assists in analysis or decision-making, that accountability does not transfer with it. It remains with the human.

When individuals are responsible for conclusions generated through systems they do not fully understand, control, or trust, quiet resistance becomes the rational response. Teams do not openly reject the technology. They simply return to familiar workflows.

Scaling AI Is Not a Technology Challenge. It’s a Workflow Design Challenge.

Workflow redesign is the most underinvested lever in AI adoption.

Layering AI tools onto existing processes rarely produces meaningful change. Organizations must map workflows honestly, remove redundant steps, and decide which tasks should be automated, simplified, or eliminated. If AI adds steps rather than removing them, adoption quickly disappears. This work is not primarily a technical challenge. It is organizational design.

Governance architecture must also be established before deployment, not after. This includes defining validation pathways, documentation standards, and clear boundaries around what AI informs, what it automates, and where human approval remains required. When these guardrails are clear, innovation moves faster because teams understand what is permitted.

Leadership alignment must be anchored in outcomes rather than aspiration. Statements such as “we are adopting AI” are not strategies. Transformation requires measurable objectives: faster development timelines, improved consistency of analysis, higher quality insights, or more efficient decision cycles.

Organizations must also address the skills gap directly. A GlobalData survey of 109 biopharma professionals found that 49 percent identified skill shortages as the leading barrier to digital transformation. These capabilities do not develop automatically. Companies need individuals who can operate across technical, scientific, regulatory, and commercial domains simultaneously.

The Real Competitive Question

The competitive question is no longer who experiments with AI. It’s who operationalizes it.

AI tools in biopharma are maturing quickly. Competitive advantage increasingly will not come from which platforms organizations purchase or which pilots they run. It will come from which organizations build the governance models, operating structures, and workflows necessary to use these technologies responsibly at scale.

Most organizations don’t lack the will to do this work. They lack the infrastructure and experience to do it well in a regulated environment where the cost of getting it wrong is high.

Technology vendors can deploy systems. Building the organizational capability to use them, and sustain that use over time, is a different kind of work entirely. At Opus Strategy, that is the work we do alongside leadership teams: translating AI ambition into the governance models, operating structures, and change programs that turn promising pilots into durable enterprise capability.