Building Trustworthy AI in South Asia: Region Ready to Rise

The event was organized by GAMP South Asia Steering Committee Members: Arnel Cabungcal, Charlie Wakeham, Chia Phei Kok, Pipiet (Manuh) Pitasari, Puja Pathak, Sarah Taylor, Vighnes Kanagaraj, Vivien Santillan and Zain Ul Abidin

When the GAMP South Asia CoP opened the virtual doors to the webinar “Building Trustworthy AI: Safety, Autonomy and Transparency,” what began as a technical deep dive quickly became something far more compelling: a story about a region eager, curious, and boldly stepping into the future of AI empowered pharmaceutical and healthcare operations.

And the evidence? The poll results—unexpected, telling, and deeply reflective of where the South Asia life sciences community stands today in the global AI maturity curve.

This wasn’t just another webinar. This was a regional pulse check. A collective declaration: “We are ready.”

A Community with a Purpose

Before a single slide appeared, Charlie Wakeham, Immediate Past Chair, GAMP Global and GAMP South Asia Steering Committee, reminded everyone why they were there: to extend GAMP understanding and expertise, and to create an open, safe, and practical platform for learning.

This isn’t a passive community. It’s knowledge sharing engine driven by practitioners who want more than compliance—they want clarity, confidence, and capability. This ethos set the tone for everything that followed.

Poll #1: Is AI Really Coming? The Audience’s Answer: It’s Already Here.

The first poll of the day asked: “Is your company planning AI implementation?”

Here’s what shook the room:

• 18 percent noted that AI is already implemented.
• 51 percent noted they are actively planning implementation in the next 1-2 years.
• The rest noted that they are not yet planning an AI implementation.

For a region historically described as “still emerging” in digital maturity, more than two thirds of the audience is either already implementing AI or actively preparing for it.

This was the first moment where the audience’s own voices reframed the narrative. AI adoption in South Asia is not a distant aspiration—it's an active transformation.

The second poll dove into the purpose behind AI usage: “Why are South Asian companies turning to AI?” The poll showed a perfectly even distribution between four possible responses. What this tells the industry is profound: South Asia is experimenting broadly. Teams are applying AI to documentation, insights, operational efficiency, and even non regulated problem solving.

This wasn’t a community wondering whether AI could help. This was a community asking: “Where can it help the most?”

Where Human Expertise Meets AI Capability: The AssistDent Story

The heart of the webinar was a compelling walkthrough of the AssistDent AI case study—an AI powered dental radiograph interpretation tool.

Williams graced the webinar by presenting key insights regarding building trustworthy AI. Williams—known to most simply as Tom—brought both deep academic grounding and his extensive industry experience. Beyond his contributions to the ISPE GAMP^® Guide: Artificial Intelligence and ISPE GAMP^® 5 Guide (Second Edition), he currently serves as CTO at Manchester Imaging, where he leads research and development in AI-based dental technologies. His ability to translate complex machine learning concepts into practical, regulated industry language set the tone for an insightful and accessible presentation.

In his session, attendees were taken behind the curtain:

• 1,500 x ray images
• Expert annotated training data
• U Net neural network architecture
• Deterministic outputs for consistent validation
• Sensitivity and specificity comparisons between AI, dentists, and human AI collaboration

The key message? AI is powerful, but humans remain essential.

Dentists interpret context. Dentists understand patient history. Dentists know what isn't in the x ray. The model supports decisions—it doesn’t replace them.

This theme echoed throughout the session: Trustworthy AI isn’t just about algorithms. It’s about relationships—between humans and machines.

Another poll question asked the attendees about their personal AI Usage. The result may surprise even the most seasoned technologists.

Almost 80 percent of attendees were already AI users in their daily work.

This was the moment that lit up the session— attendees realized they were not alone. They weren’t outliers. They were part of a growing, unstoppable movement.

GAMP South Asia had created a space where that shared momentum finally became visible.

Trustworthiness Principles: More Than Buzzwords

The webinar also discussed generative AI and how the large language models (LLMs) are trained on massive datasets and refined through human aligned reinforcement learning, resulting in flexible tools capable of generating clinical style text, reports, or explanations.

It was emphasized that such incredibly powerful, generative AI models require new approaches to specialization, validation, and oversight—from techniques like retrieval augmented generation (RAG) to domain specific fine tuning and model to model auditing.

In his session, Williams emphasised the broader trustworthy AI principles highlighted in ISPE GAMP^® Guide: Artificial Intelligence:

• Human autonomy and control
• Safety and security
• Fairness and bias mitigation
• Transparency
• Accountability
• Privacy and data protection
• Sustainability

These weren’t abstract concepts. Each tied directly to real validation challenges:

• How do teams test for bias?
• Where do teams enforce transparency?
• How do teams balance human oversight with automation?

These questions became the cliffhangers of the day—answered not immediately but reserved intentionally for a GAMP South Asia CoP March Q&A session, exclusive to CoP members, where Williams will return to take the conversation even deeper.

For the final poll of the day, the attendees were asked if they are using publicly available AI models or enterprise models. The outcome: South Asian respondents are using hybrid approaches.

The insights here are as strategic as they are practical:

• Public models dominate early experimentation
• Hybrid usage is rapidly growing
• Enterprise training is still rare—but rising

The takeaway? South Asia is learning fast, exploring boldly, and adapting in real time. And that’s exactly why this webinar mattered.

A Webinar That Sparked Dialogue, Not Just Learning

At the end of the session, attendees were asked one final open-ended question: “What is your key takeaway?”

Responses poured in:

• The need for transparency
• The importance of choosing AI model wisely and human in the loop systems
• Recognizing risk based validation as a must
• Recognizing that LLMs are deliberately created as non-deterministic, it is not a design flaw!
• Awareness of hidden bias and model thresholds
• Curiosity about regulatory detection of bias
• Questions on autonomy vs oversight

As one of the GAMP South Asia Steering Committee members, Zain Ul Abidin, Deputy Director, Drug Regulatory Authority of Pakistan, summarized: “Trustworthy AI is not about using the biggest or newest model. It is about choosing the right tool for the right purpose. In regulated environments, transparency, safety, and clear human oversight matter more than technical sophistication. Strong governance and understanding of limitations are key to using AI with confidence.”

The conversation didn’t end there. It has expanded—spilling into follow up questions, future topic ideation, and conversations planned for the March member only session.

This is the power of a community that cares.

Why This Webinar Mattered—and What’s Next

What truly made this event special wasn’t only the expertise, the use case, or the frameworks. It was the connection—the realization among attendees that:

• They are not behind.
• They are not alone.
• AI adoption in South Asia is accelerating—intelligently, responsibly, and collaboratively.

And central to that acceleration is the GAMP South Asia CoP—not just with hosting webinars, but in cultivating a regional AI ecosystem grounded in safety, autonomy, transparency, and good practice.

And judging from the polls, the questions, and the momentum, this is only the beginning.

Final Thoughts

This webinar didn’t just teach the region about trustworthy AI. It showed the region it is ready for trustworthy AI.

The South Asia CoP continues to transform curiosity into capability, questions into clarity, and individuals into a vibrant community.