Toward Certified AI: Deploying Generative Systems Within Federal Risk Frameworks (M11c)
As generative AI adoption accelerates, federal agencies face the challenge of deploying Large Language Models (LLMs) in environments that demand auditability, compliance, and trust. Traditional certification and software assurance frameworks are ill-equipped to address the unique risks posed by dynamic, emergent AI systems. This talk delivers a practical blueprint for deploying compliant, secure-by-design LLM applications: chatbots, copilots, and RAG pipelines, aligned with the NIST AI Risk Management Framework and Executive Order 14110. From prompt injection and hallucination to model supply chain risks, the talk maps critical threats to policy mandates and introduces guardrails for real-world implementation. Attendees will walk away with actionable strategies to align LLM deployments with Zero Trust architecture, build AI-specific governance workflows, and prepare for the evolving landscape of AI certification and oversight.