Discovering Bias in Large Language Models (LLMs)

Responsible Tech Forum
ACM Distinguished Speaker Series 2026
Date: Friday, September 25 | Time: 12:00 PM – 1:00 PM CT
Format: Hybrid - Virtual + Cypress, TX
Session Format: Distinguished Lecture + Interactive Discussion

In-person location details will be provided to registered attendees.

***

Discovering Bias in Large Language Models (LLMs)
Speaker: Mehdi Bahrami – ACM Distinguished Speaker

***

Overview
As large language models become increasingly embedded in enterprise systems, healthcare, education, decision support, and everyday digital tools, understanding their capabilities is no longer enough. We also need to understand how these systems can reproduce bias, reinforce stereotypes, generate misleading information, and create unintended consequences when deployed in the real world.

This ACM Distinguished Lecture will explore how bias emerges in large language models, how it manifests in model outputs, and why it matters for the people building, deploying, governing, and using AI systems.

Participants will examine both the technical foundations and broader implications of bias in generative AI, including approaches for identifying, evaluating, and mitigating these risks.
Through practical examples and research-informed approaches, the session will advance the conversation from simply recognizing AI bias to understanding what developers, organizations, researchers, and technology leaders can do about it.

***

What You Will Learn

• How bias can enter and manifest in large language models
• Common forms of bias, stereotypes, hallucinations, and misinformation in LLM outputs
• Methods and tools used to evaluate and detect bias in generative AI systems
• Emerging approaches for mitigating bias and improving fairness
• Practical considerations for deploying more trustworthy and accountable AI systems

***

Why This Matters
Organizations are moving quickly from experimenting with generative AI to integrating AI into products, workflows, decision-making, and increasingly autonomous systems.

As adoption accelerates, questions of fairness, transparency, evaluation, accountability, and human oversight become operational concerns and not simply ethical principles.

Understanding where bias comes from and how to evaluate it is therefore becoming an important capability for anyone responsible for building, deploying, purchasing, governing, or using AI systems.

This session supports the Responsible Tech Forum's mission to help practitioners move responsible technology from principles to practice by strengthening the community's ability to build, evaluate, and govern trustworthy intelligent systems.

***

Who Should Attend

• AI engineers and developers
• Data scientists and machine learning practitioners
• Product managers and technology leaders
• AI governance, risk, and responsible AI professionals
• Researchers, students, and academics
• Anyone interested in ethical, responsible, and trustworthy AI

***

Part of the ACM Distinguished Speaker Series hosted by Responsible Tech Forum.

Registration is required. Virtual access and in-person participation details will be provided through Meetup.