AI Risk – Can We Balance Innovation and Security? and The Expert Paradox: Engineering the Boundaries Between Probabilistic AI and Deterministic Reality (Possibly merging 2 topics)

This presentation is designed to provide participants with a foundational understanding of the complexities and evolving challenges of AI risk management. The learning objectives focus on equipping attendees with the knowledge to navigate the rapidly changing landscape of artificial intelligence, where historical precedents are limited and guidelines can quickly become outdated. Realistically speaking, between the due date for presentations and the actual presentation itself, the discussion can evolve.

By the end of the session, participants will be able to:

• Identify and address the unique limitations of AI technologies, particularly the probabilistic nature of AI models and the need for repeatable, well-documented AI-assisted design processes. They will also explore the implications of AI-enabled automation and how it impacts organizational workflows.
• Grasp the dynamic tension between traditional IT governance structures and the urgent drive for rapid AI adoption. The session will address balancing these concerns, including the risks of not engaging with frontier AI models, why interacting with models in 2026 may differ significantly from previous experiences, and why your users want to use their personally customized models versus enterprise needs like logging, price stability, and data leakage.
• Understand additional complexities of enterprise sourcing in the AI domain, including the necessity of selecting and integrating different AI models for diverse tasks, and why AI-to-AI communication makes everything worse from a security perspective, using examples like MCP servers, Moltbook, and Alphastar. By articulating these challenges and solutions, the presentation aims to prepare attendees to manage AI risks effectively and responsibly within their organizations.

The prevailing venture capital narrative claims that enterprise AI is a headcount-reduction tool designed to displace labor. The ground-level reality facing A/E/C firms is exactly the opposite: AI commoditizes yesterday’s baseline competence and immediately inflates demand for high-level expert judgment. This interactive session breaks down the “frame problem” in generative AI and explores why language models require human practitioners to define, constrain, and validate their outputs. We will examine the operational friction between probabilistic AI models and the deterministic accuracy required for stamped engineering deliverables. Moving beyond the hype of autonomous workflows, this presentation provides a factual, field-tested blueprint for cultivating hybrid professionals—translators who bridge the gap between computer science and engineering liability—and scaling firm volume without compromising standard of care.

Earn 1 PDH