The Connection is a monthly briefing for pharmaceutical supply chain insiders to cut through the noise and surface clear, opinionated insights on data, compliance, and the systems that run the industry.
FDA Consolidates Data Systems for AI / FDA
What happened
The FDA announced a major expansion of its internal AI capabilities with the rollout of Elsa 4.0 alongside a new consolidated data platform called HALO.
The initiative combines more than 40 separate FDA application, submission, and operational systems into a centralized environment designed to support AI-driven workflows across the agency.
Why it matters
This is bigger than "the FDA is using AI."
The real signal is infrastructure consolidation. The agency is reducing fragmentation between systems so AI can operate closer to the actual operational layer of regulatory work.
Organizations with the cleanest, most accessible, and most connected data will move faster as regulatory workflows become increasingly machine-assisted.
In practice
The interesting part here isn't document generation or chat interfaces. It's the integration layer underneath them.
The FDA explicitly stated that staff previously had to bring data to Elsa, but now Elsa sits on top of FDA systems directly.
That's the same transition the industry is beginning to face. AI becomes far more useful once systems, submission data, serialization data, quality systems, and operational workflows are connected instead of siloed.
If your data is hard to move, your system is hard to evolve.
More on this signal:
Read the FDA announcement
Big Pharma Buys Time / Reality Check
What happened
Biotech dealmaking is on pace for a major year, with Big Pharma accelerating acquisitions to strengthen pipelines ahead of major patent expirations. Q1 biotech M&A reached $84 billion, nearly double the $44.4 billion recorded a year earlier.
Why it matters
This is what the patent cliff looks like before it shows up in revenue.
Large manufacturers are not just buying promising science. They are buying time, optionality, and future revenue coverage before exclusivity losses hit key products.
More than $300 billion in sector revenue faces loss of exclusivity over the next five years. That kind of pressure changes behavior across the supply chain, not just in boardrooms.
In practice
Every acquisition creates downstream operational work.
New products, new sites, new partners, new packaging configurations, new serialization responsibilities, and new data flows all have to be absorbed into existing systems.
That is the quiet part of pharma M&A. The deal gets announced at the corporate level, but the integration work lands in regulatory, quality, supply chain, and data operations.
If the industry is entering a heavier M&A cycle, companies need systems that can absorb change without rebuilding every connection from scratch.
More on this signal:
Read the Reuters Story
GS1 Digital Link / Resolver Reality
What happened
I sat down with Patrick Maher for his perspective on GS1 Digital Link, compliant resolver technology, and how these concepts fit into pharmaceutical supply chain.
"Digital Link doesn't create value by itself," he explained. "It provides the standardized foundation that allows value to be deployed consistently, governed centrally, and scaled without operational chaos."
Interview
Q: For pharma supply chain leaders evaluating Digital Link, where should they start?
Patrick: "The starting point should always be: What outcomes do we need to deliver - and to whom?" Digital Link is "a standardized way of structuring identifiers as web-native URIs" that allows the same product code to connect different stakeholders to different governed services and information.
Q: How does this interact with DSCSA, EPCIS, and serialization?
Patrick: "They don't compete - they complement each other." He described EPCIS as the event layer, DSCSA and serialization as the compliance and status layer, and Digital Link as "a standardized entry point into that landscape, typically via a resolver."
Q: What is the biggest adoption challenge?
Patrick: "The biggest barrier to meaningful Digital Link adoption in pharma is not the standard itself, but the ongoing governance and maintenance of resolvers." He pointed specifically to concerns around ownership, availability, auditability, and accountability for regulated information flows.
Why it matters
Digital Link is often discussed as if the QR code itself is the innovation. Patrick points out that the real complexity sits underneath: resolver logic, backend integrations, governance rules, and the quality of the underlying data.
"The hard work is in data models, ownership, quality, and governance."
In practice
A scan is only as trustworthy as the system it resolves into.
If the underlying product, regulatory, sustainability, or supply chain data is incomplete or poorly governed, Digital Link will not fix the problem. As Patrick put it, "a beautifully implemented Digital Link won't fix that; it will just expose the weaknesses faster and more publicly."
More on this signal:
Connect with Patrick Maher on LinkedIn
Shadow AI in Pharma / State of AI in Business
What happened
Shadow AI is spreading across biopharma R&D as scientists use public AI tools to interpret results, refine protocols, and structure experimental thinking outside approved systems. The article points to research showing only 5% of scientists can analyze experimental results independently within official tools, while 77% report using public AI tools in lab work.
Why it matters
This is not just an AI governance story.
It is a systems story. Shadow AI grows when official tools capture data but fail to support interpretation, comparison, and decision-making at the pace employees need.
The article makes the stronger point: shadow AI is infrastructure feedback. When approved platforms make useful work harder, people route around them. That creates visibility, auditability, and data integrity problems.
In practice
A system can be compliant and still be operationally weak.
The fix is not just tighter policy. It is better architecture: approved systems that bring AI into governed workflows instead of forcing scientists to choose between speed and control.
If the useful work happens outside the system of record, the system of record is no longer where the real work lives.
More on this signal:
Read the Medicine Maker story
Manufacturing Excellence / Connected Systems
What happened
Manufacturers still don't operate true global production systems, despite years of investment in digital transformation and operational excellence programs. That is what McKinsey is saying in a survey of manufacturing COOs: 74% said they had a global production system, but only 29% said it was fully implemented across their organizations.
None reported fully integrating AI into operational decision-making.
Why it matters
A lot of organizations have standards, tools, dashboards, and transformation programs. Fewer have systems where data, workflows, decision-making, and operational learning actually connect across the network.
A production system becomes real when insights continuously translate into frontline action and learning loops compound across sites.
That is the difference between digital infrastructure and operational intelligence.
In practice
Operational risk is fragmentation disguised as modernization.
Disconnected pilots, partially adopted systems, isolated AI initiatives, and inconsistent site practices create environments where improvements appear locally but fail to scale network-wide.
The real competitive layer in manufacturing and supply chain operations is not who has the most software, but who can turn operational data into repeatable execution across sites, partners, and workflows.
More on this signal:
Read the McKinsey article
Speaker Spotlight / John Winkler
What they're doing
John Winkler, CTO of Trust.med, will be speaking at GS1 Connect 2026 in Las Vegas during the session 607: Interoperability Meets Accountability: Scaling EPCIS Connectivity With Conformance and Control.
Why it stands out
The session takes place Thursday, June 11, 2026, from 1:30 PM - 2:15 PM PT and will explore how open, interoperable EPCIS networks and continuous data validation can improve onboarding, reduce repeat errors, and strengthen digital supply chain resilience.
In practice
John will be joined by Gateway Checker, with the session focused on a practical question for the post-DSCSA supply chain: how do we move from simply exchanging EPCIS data to scaling connectivity and continuously testing and improving the quality of that data?
More on this signal:
View the GS1 Connect agenda
Final / Thought
Across manufacturing, traceability, and Digital Link the same pattern keeps showing up.
The bottleneck isn't intelligence.
It's infrastructure.
Disconnected systems, poor data quality, and operational fragmentation aren't disappearing. The organizations that will move fast while being prepared for the future will be those with systems capable of turning data into trusted, connected execution.
Have a story to share?
Send me an email
Want to learn more about Trust.med?
Visit our website
Interested in becoming a tech partner?
Book a meeting
More from Trust.med
Keep exploring supply chain connectivity.
Read more practical notes on partner onboarding, data exchange, and network operations.