Pre-Conference Course on Basics of AI and 2nd Annex 22 / AI Conference 2026

Programm

• History of AI
• Types of AI
• Current situation and real-life examples
• Technological Basics
• Different learning / training Methods
• Example use cases

• Overview of core model architectures and data types
• Data splitting: training, validation, and testing phases
• Ensuring data quality, representativeness, and overcoming common bottlenecks

• Specialized vs. General-Purpose AI: Understanding the key differences
• Step-by-step training process, using cancer diagnostics as a case study
• The roles of pre-training and fine-tuning
• Ensuring AI traceability and explainability in critical Tasks

• Fundamentals of Large Language Models (LLMs)
• Strengths and limitations of LLMs compared to specialized AI
• Introduction to LLMs acting as autonomous agent

• Applying ICH Q9 QRM to AI/ML
• QRM throughout the System Life Cycle
• Thinking critically about AI-enabled Computerized Systems
• Using the AI Maturity Model
• Achieving Data and Model Governance

• Challenges facing the life sciences industry in dealing with AI and ML
• GAMP® 5 meets AI: The GAMP AI Guide
• Use cases for AI in pharma and biotechGMLP (Good Machine Learning Practice): An SDLC for AI/ML

• General introduction
• Brief introduction to AI & ML
• Drivers for using AI & ML in pharma
• Regulations and guidance

• Current situation
• What do you need to watch out for?
• What are the risks?

• Human in the loop requirements
• Supplier overview & collaboration
• Long-term sustainability
• Dynamic systems / continuous learning
• Do AI systems, particularly (autonomous) agent systems, require a paradigm shift in validation (Validation 4.0)?

• Open-Source vs. Closed-Source: Transparency, control, and dependencies
• Bias in training data and its impact on results
• Data privacy and confidentiality when using generative AI
• Hallucinations: Why LLMs generate convincingly false information

• Regulatory requirements (EU AI Act, Annex 11, 22, Chapter 4, Part 11).
• AI governance: managing risks associated with non-GxP systems in a GxP environment.
• Validation: CSV as a holistic approach for compliant AI.
• Human-machine interaction: Human-in-the-loop vs. human-centric concept.
• Practical examples: Benefits of AI in quality assurance / Thoughts on the potential and risks of AI for patient care
• Outlook: The future role of AI agents.

• Why QA cannot stay on the Side-lines?
• What makes AI different from traditional computerised systems?
• Practical quality challenges observed in projects
• Inspection readiness: what QA should already prepare for?

• What do organizations need to consider when approaching AI in GxP
• How does AI strategy interconnect within a typical corporate strategy layout?
• What governance functions are relevant and how do they integrate
• How can enabling elements like data, Quality, and technology be activated

• Why AI struggles in GxP environments: The reality behind hype, pilots, and Shadow AI
• Regulatory expectations for AI: Trust, validation, explainability, and compliance
• Building the foundations for trusted AI: Governance, data readiness, and organisational capabilities
• From foundations to governed AI: Practical frameworks and maturity models
• When AI becomes a GxP system: Validation, risk control, and leadership responsibilities
• Beyond traditional validation: Patient-facing LLMs, adversarial testing, and continuous validation
• Leadership takeaways: How organisations move from experimentation to trusted AI

• Platform mindset: Embracing a platform-oriented approach to leverage GenAI capabilities effectively.
• Democratization of Technology: Ensuring easy access to GenAI tools across the enter-prise, empowering end-users to innovate responsibly.
• Quality Risk Management: Implementing a comprehensive quality risk management framework to evaluate and mitigate potential quality impacts associated with GenAI usage.

• Maintenance Challenges
• How knowledge databases work
• Framework Conditions for the Use of the AI-Based Assistant
• Quality Control of Knowledge Entries
• Benefits, Impact, and Outlook

• Overview: State-of-the-art cyber security & resilience
• AI as a threat versus AI as an opportunity
• Examples AI-based attacks
• Urgent call to action for the industry

• Critical Reflection on AI in Maintenance
• Expectations and roadblocks
• Limits of data driven Models
• Pragmatic alternatives
• Practical experience at CSL

• Basics for QC & QA on IT framework - digitalization - automation - use of AI
• Generation of raw data and data systems
• Real-life automation examples and AI examples from QA & QC