AI Errors May Be Impossible to Eliminate

What That Means for Its Use in the FDA

Examine how organizations can evaluate, govern, monitor, and justify AI-supported decisions when system outputs may remain inherently variable within regulated operations.

EXCLUSIVE: Get your team to

attend for just $ 290.

Individual registrations available too.

COURSE OVERVIEW & STRUCTURE

AI systems introduce different control challenges than traditional software.

Organizations are deciding where AI can support regulated activities, where human review remains necessary, and how accountability should be assigned. This session organizes the practical considerations surrounding AI variability, validation boundaries, oversight expectations, monitoring approaches, and governance responsibilities in GMP environments.

The discussion is structured around the characteristics of probabilistic systems, regulatory expectations, validation limitations, governance responsibilities, performance monitoring, and operational use cases. Examples from quality, manufacturing, compliance, and inspection settings are used to examine how organizations maintain control when AI outputs are not always predictable.

WHAT YOU’LL LEARN

Managing AI Systems That Will Occasionally Be Wrong

Identifying weaknesses hidden by checklist-driven audit approaches
Building risk-based audit plans using operational performance data
Using deviations, CAPAs, complaints, and trends as audit inputs
Applying investigator-style “show me” questioning techniques
Tracing records and evidence through actual process execution
Recognizing red-flag indicators inspectors identify immediately
Converting audit observations into preventive system improvements

WHO SHOULD ATTEND

Designed for professionals evaluating, governing, validating, monitoring, or overseeing AI-supported activities within regulated operations.

Quality Assurance

Validation

Compliance

Regulatory Affairs

Data Governance

Operations

Manufacturing

Risk Management

FACULTY

Dr. Ginette Collazo

Dr. Ginette Collazo combines expertise in human reliability, human error reduction, and regulated industry operations, making her uniquely positioned to address AI oversight challenges. Her work has focused on understanding how people interact with complex systems, investigate failures, establish effective controls, and improve decision quality. This perspective provides practical insight into managing AI-supported processes while maintaining accountability, compliance, and operational control.

AFTER YOU REGISTER

After you register your seat, our team organizes everything for you; from access to certification – “from one place”.

You receive a confirmation and your course appears instantly in your TalkFDA dashboard.
Your course page becomes your central hub where you can join the session, access materials, and manage your learning.
At the scheduled time, simply click Join Session and you’ll be connected to the live Webex session with the instructor.
After the session, the playback becomes available on the same course page so you can revisit important sections.
Complete the course and your certificate is unlocked automatically in your learning history.
Everything related to your course: access, materials, playback, and certification – lives in one place.

For Teams Seeking Deeper Implementation Support

This topic can also be delivered as a customized session through TalkFDA Elite Training Labs for organizations seeking structured, implementation-focused training aligned with their internal processes and quality systems.

Elite Training Labs sessions are tailored for teams and delivered in formats designed to support application, discussion, and practical integration beyond the live webinar format.

◙

FDB1304

🏠︎