2026 ISPE AI in Life Sciences Summit Recap

ISPE has hosted a summertime conference in Boston for the last three years (2021 was virtual, and in 2023 it was held in Dublin). The conference has remained a summer event focused on biotech and life-science solutions. This year the Boston conference was rebranded as the ISPE AI in Life Sciences Summit, powered by GAMP. Some clear differences emerged:

1. Attendee breakdown. It was clearly a different flavor of attendees this year. Technology and quality professionals were front and center, with many use cases focused on what pharma companies are building and doing to optimize their processes.
2. A clear chasm of maturity. Some sessions focus on management workflows and what can be optimized, while others dive into engineering-related processes that can be improved. Overall the former had more presence and more mature case studies.
3. Clear maturity around technology, its purpose, and the diagnosis behind it. A few topics and keywords stood out:
   • Shadow AI: users adopting technology regardless of their organization's procurement of that technology.
   • Data maturity and presence: proper segmentation and management of data.

Other areas of interest:

• AI rhetoric and understanding have improved a lot in this conference forum. I'm not sure the actual presentation quality has moved with it.
• Lots of talk about AI and use cases, but pointed, concrete case-study examples were rarely clear.
• Data, data integrity and security, and on-prem use and management came up throughout. Clearer here than at other events: there's a real disconnect between smaller-org vendors and their rhetoric versus the quality-focused, deeply-backgrounded professionals they're selling to. That gap is obvious.

Day-by-Day Breakdown

Day 1 - June 22, 2026

Opening Remarks: Welcome to the Summit
Mike Martin, ISPE

• Shadow AI: people use the best tools with or without org support.
• You only innovate as fast as your org's tooling allows.
• Great remarks on finding people to collaborate with. Try to ask others challenging questions, at least one a day, and get dialogue going.

From Lab to Patient: Governing the AI-Driven GxP Pipeline
Bill Reid, Google

• Strong governance creates the confidence to innovate at speed. He likened it to the brakes on an F1 car: the brakes are what let it go fast. It is the difference between just letting off the gas and using both the brakes and the gas.
• I had the opportunity to sit down with Bill Reid after this session. Two clear points he made:
• Data integrity is critical, and not just in the sense that one is aware of it. The future of how data gets properly controlled, stored, password-locked, and provisioned is becoming increasingly concerning.
• Bill also remarked on SAIF, the Secure AI Framework.

Regulatory Keynote: Artificial Intelligence, Critical Thinking, and the Enduring Principles of Quality
LT Seneca Toms, FDA

• Decades of data generated, the gap is the knowledge and understanding of it.
• "Decision support tools work best when combined with AI."
• The other impressive part of this talk was not just its brevity and clarity, but its openness. It was apparent to me that Seneca did not use AI-generated slides or visuals, something some presenters had that felt both unnatural and information-dense. Having clear, concise slides and points to share is still deeply valuable, and well done in this case.

Conference Highlights
Nick Armstrong, IPS

• AI as collective intelligence.
• "You're not just attending this conference, you're contributing to the guidance it creates."
• Named the big pharma in the room as the consensus. Referenced FDA and MHRA.

Note: these sessions are not exhaustive. There were two tracks, so I only sat in on one, and I did not attend every session.

Scaling GenAI: A 5-Year Roadmap for Global Biologics
Jesse Summers, AstraZeneca

• Roadmap for company-wide AI adoption and education.
• My questions: how do you get teams to want AI education? The 97/3% stat implies ~100 builder projects. Have you run a hackathon? What's happening in inspection intel?
• Jesse shared two real takeaways: (1) the company has held a hackathon for builders and those using the new tools they are procuring, and (2) most of what is being built or tested doesn't need a large amount of oversight until a later implementation phase at a larger enterprise level.
• I found some challenge to this approach, but it's clear it can work really well for management-related activities and is a great way to get the workforce involved.

Guardrails for Building Pharma-Trusted AI in Manufacturing
Rick Johnston, PhD, Modersys

• Showed flaws in a Gen AI image via a prompt.
• "We're going to put a chatbot on top of our data." Demoed handling a manufacturing issue.

AI in Regulatory Affairs and CMC: From Digital Tools to Submission-Ready Governance
Kelsey Hoontis, MSc

• Talked about embedding practical AI and regulatory tools for the preparation and structuring of data and submissions.

AI in Regulatory Intelligence and Global Compliance
Teginder Singh, Google

• Talked about the Google reg-intel approach to monitoring.

Risk-Based AI in Practice: Compliance and Quality
Mike Salem, Gilead, and Alison Sathe, SVP of Quality Intelligence at Redica Systems

• A well-grounded, practical approach across several topics:
  • Risk scoring.
  • Risk-based audit prioritization, risk-ranked portfolio, site risk profiles, global enforcement pattern mapping.
  • Guidance references: Health Canada, TGA, and the FDA credibility assessment. ELSA came up.
  • All really good ways to articulate what we do.
• I was able to ask Mike about the best ways to disseminate this data within the organization. He noted you want a data scientist or SME in true control.

Day 1 Final Remarks

Great benchmarking for where industry players are at, and helpful to see where vendors and other SMEs sit in terms of expertise, literacy, and ability to conduct themselves.

Day 2 - June 23, 2026

Day 2 Opening Remarks
Mike Martin, ISPE — wrapped ~9:14.

Unlocking AI in Regulated Industries: Shifting Mindset
Marcel Geers, Porsche Consulting

A Look into Industry's Innovative Digital Regulatory Transformation
Michael Abernathy, AVP, Amgen

• Co-developer of HL7 FHIR® (Fast Healthcare Interoperability Resources, pronounced "fire"), the standard built to improve product-quality and CMC data exchange.
• Biggest takeaway: the life and cost burden of the regulatory ecosystem.
• "We have an obligation to give parents more tomorrows." Dori Gonzalez joined him on stage to discuss the impact of chronic conditions and our need to drive innovation against the regulatory bottleneck.

From Generative AI to Validation Feasibility
Gourav Pandey, Takeda

• Focused on longstanding challenges around fragmented insights such as compliance data, SOP changes, and CAPA changes, and the silos the data sits within.

Practical Guide to Measuring LLMs
Nikolai Makaranka, CEO and Founder, Daikon

• Measuring and understanding the ability of the technology to write and polish SOPs, and frameworks like Annex 22 guidance, which significantly limits the use of modern tech. He gave a great history of how AI has been banned throughout time.

Nikolai Makaranka, CEO and founder of Daikon.

AI in Regulated MedTech: What Regulators Expect and Where Manufacturers Fall Short
Jennifer Bromm and Xia Huang, DEKRA

• Med device and pharma colliding as drugs need delivery devices.
• KeyPedia metaphor: a product is like a child, be proactive, not reactive. KeyPedia is the owner's manual and safety net.

De Novo Design of a Biased GLP-1R Agonist Miniprotein
Chris Bahl, PhD, AI Proteins

• "The data layer of quality decisions."

Brandi Stockton (left) and George Kwiecinski in front of the ISPE Life Sciences poster on Day 2.