​Guardrails for Gen AI: From Policy to Practice

How to constrain AI systems without killing their value

​Everyone uses the word "guardrails." Almost nobody means the same thing.

​This co-presented session brings together AI strategy and cybersecurity expertise to break down what guardrails actually are, where they belong in your AI systems, and how to implement them without crippling the tools your teams rely on. We will cover the strategic layer — governance, risk classification, and authority boundaries — then dive into the technical layer with hands-on walkthroughs of real guardrails frameworks like NVIDIA NeMo Guardrails and Guardrails AI. Whether you are a business leader trying to deploy AI safely or a security practitioner trying to contain it, this session gives you a practical, layered approach you can take back to your organization immediately.

​What you'll learn

• ​Why traditional security assumptions break down with AI systems

• ​The three layers of AI guardrails: governance, architecture, and implementation

• ​How to treat AI agents as high-risk identities using least privilege and Zero Trust principles

• ​Where guardrails actually get enforced — at system boundaries, not in prompts

• ​Real tools and frameworks for building guardrails into AI pipelines

• ​What guardrails can and cannot do — honest limitations and tradeoffs

• ​Hands-on homework: get started with Guardrails AI and NeMo Guardrails in 15 minutes

​This is not about adding a system prompt and hoping for the best. It is about building AI systems that fail safely and stay auditable.

​Your speakers

​Andrew Stafford is a Fractional Chief AI Officer and Director of Technology and Global Alliances at The AI Collective. He works with executive teams to establish clear ownership of AI decisions, focus on real ROI, and avoid unnecessary risk, tool sprawl, and hype-driven strategy. Andrew brings 25 years in tech spanning Amazon, NASA, and five startups to the strategic and architectural side of AI security.

​Ken is a security practitioner with ISSA Hampton Roads. He brings deep hands-on experience in security engineering and implementation, and will lead the technical deep dive into guardrails tooling, configuration, and real-world limitations.
Together, Andrew and Ken will walk you through a complete guardrails framework — from policy to architecture to working code — blending strategic thinking with practical implementation.

​Who should attend?

• ​CISOs, security architects, and IT leaders responsible for AI risk

• ​Developers and engineers building or deploying AI agents and applications

• ​Product managers and business leaders evaluating AI tools for their teams

• ​Security practitioners curious about the AI threat model

• ​Anyone responsible for AI governance, compliance, or acceptable use policies

• ​Technical professionals who want hands-on experience with guardrails tools

​No prior AI security experience required, but familiarity with basic AI concepts is helpful.

​Why this matters

​AI systems collapse data, logic, and control — and traditional security was never built for that. The real danger is not bad AI. It is trusting non-deterministic systems with authority. Organizations that get guardrails right will deploy AI faster and safer. Organizations that skip them will learn the hard way. This session is about making sure you are in the first group.

​FAQs

• ​​​​Is there free parking in the area?
  Yes. There's lots of parking on the ECPI University campus

• ​​​​Will anything else be provided?
  Yes, snacks & drinks will be provided by our awesome partner, TEKsystems & Armon Wade

• ​Is this free?
  100%, but snag your spot soon!

​​🤝 Organizers & partners

​​Organizers

​The AI Collective Hampton Roads

​​Platinum partners

​Bronze partners

​​Our parent org

​The AI Collective is a global non-profit community uniting 150,000+ pioneers – founders, researchers, operators, and investors – exploring the frontier of AI in major tech hubs worldwide. Through events, workshops, and community-led research, we empower the AI ecosystem to collaboratively steer AI’s future toward trust, openness, and human flourishing.

​All attendees and organizers at events affiliated with The AI Collective agree to our privacy policy and are subject to our code of conduct. Additionally, by registering for this event, you agree to The AI Collective Hampton Roads chapter photo & video policy.