Table of contents
At LogiPharma 2026, one theme came up in nearly every conversation we had: AI.
On stage and across the show floor, the narrative was consistent—autonomous supply chains, intelligent agents, predictive orchestration.
But in smaller, off-stage conversations with supply chain leaders, a more nuanced picture emerged.
- Planners still reconciling spreadsheets.
- Procurement teams following up over email.
- Forecasts and inventory data shared in PDFs and offline files.
- ERP transformations stretching across years.
There’s real momentum behind AI. But there’s also a growing recognition that something more foundational needs to come first.
The Gap Between AI Ambition and Operational Reality
One comment we heard repeatedly captured it well:
“We need to digitalize our data and processes end-to-end before we can deploy AI agents.”
Many organizations have already begun experimenting with AI. But scaling it into real, day-to-day execution—especially across partners—remains difficult.
Part of the challenge is structural. Supply chain data still lives in different systems, formats, and timelines across organizations. Even within a single company, it can be hard to create a consistent, real-time view.
AI depends on that consistency. Without it, even the most advanced models have limited ability to act in meaningful ways.
The Invisible Layer: Manual Coordination
Another theme that surfaced often was the amount of manual work still required to keep operations running.
As one supply chain leader put it:
“There is a lot of manual work to coordinate supply and demand planning and manage disruptions with CMOs. We are often forced to adapt to the CMO’s way of working.”
Across many teams, coordination still happens through a mix of emails, spreadsheets, portals, and point solutions. Transactions move, but they don’t always align in real time. Context gets lost. Exceptions take time to resolve.
In that environment, it’s hard for AI to do more than assist at the margins.
Agents rely on timely, structured data to make decisions. When information is delayed or manually updated, the gap between what’s happening and what the system “knows” becomes difficult to close.
Why ERP Progress Doesn’t Automatically Translate to AI Readiness
ERP transformation is another major focus area across the industry right now.
One executive described it this way:
“Our ERP migration has taken two and a half years and still isn’t done—it’s a nightmare.”
These programs are critical, and many organizations are investing heavily in them. At the same time, they’re primarily designed to improve processes within the enterprise.
What came through in our discussions is that AI—particularly agent-based models—depends just as much on what happens between enterprises.
When each partner operates on different systems, timelines, and integration models, coordination becomes harder to scale. And every system change—whether internal or external—can introduce new complexity.
Some leaders are starting to explore alternatives that make it easier to integrate, exchange contextual business transactions, and collaborate on shared business processes across partners without rebuilding integrations each time. It’s an area that’s getting more attention as AI initiatives move from pilots toward broader deployment.
Complexity Is Compounding—Especially with CMOs
The increasing reliance on CMOs is adding another layer of complexity.
As one director shared:
“We’re onboarding one new CMO per year and cannot sustain that growth with our current model.”
Each new partner brings different systems, formats, and ways of working. Forecasts may be exchanged manually. Visibility into production and inventory can vary. Onboarding can take time.
Individually, these challenges are manageable. Collectively, they start to compound.
And because AI depends on consistent, real-time inputs across all of those relationships, the variability makes it harder to apply at scale.
Confidence in AI Is Closely Tied to the Foundation Beneath It
Interestingly, skepticism about AI wasn’t really about the technology itself.
It was more about the environment it operates in.
Many leaders pointed to the need for stronger data foundations, better visibility across partners, and clearer governance before expanding AI into more critical processes.
In regulated supply chains, that makes sense. Decisions need to be traceable, auditable, and reliable—not just fast.
There’s growing interest in finding ways to balance autonomy with control—using AI to augment decision-making while ensuring the right guardrails are in place.
What TraceLink is Hearing—and What It May Point To
Stepping back, the conversations at LogiPharma didn’t suggest a lack of enthusiasm for AI. If anything, the opposite.
But they did highlight a shift in focus.
From what AI could do…to what needs to be true for AI to work within the life sciences operating environment.
That shift is bringing more attention to the underlying digital infrastructure of the supply chain—how data is shared, how partners link, and how processes are coordinated across organizations.
The companies making progress here aren’t necessarily the ones with the most advanced models. They’re the ones working to make their data more accessible, their processes more connected, and their partner networks easier to operate within.
Moving Forward
AI will almost certainly play a major role in the future of pharmaceutical supply chains.
What LogiPharma 2026 made clear is that realizing that potential is as much an operational challenge as it is a technological one.
For many organizations, the next step isn’t just experimenting with new AI capabilities. It’s continuing to strengthen the digital and collaborative foundation those capabilities depend on.
That’s where much of the focus—and opportunity—seems to be heading.
