Thank you for sending your enquiry! One of our team members will contact you shortly.
Thank you for sending your booking! One of our team members will contact you shortly.
Duration 21 hours
Course Outline
Foundations: Threat Modeling for Agentic AI
- Categories of agentic threats: misuse, escalation, data leakage, and supply-chain vulnerabilities
- Adversary profiles and specific attacker capabilities targeting autonomous agents
- Identifying assets, trust boundaries, and critical control points for agent operations
Governance, Policy, and Risk Management
- Governance frameworks for agentic systems, including roles, responsibilities, and approval gates
- Policy formulation: acceptable use, escalation protocols, data handling, and auditability
- Compliance requirements and strategies for evidence collection during audits
Non-Human Identity & Authentication for Agents
- Constructing agent identities: service accounts, JWTs, and short-lived credentials
- Least-privilege access models and just-in-time credentialing practices
- Managing identity lifecycle, rotation, delegation, and revocation
Access Controls, Secrets, and Data Protection
- Implementing fine-grained access control and capability-based patterns for agents
- Secrets management, encryption in transit and at rest, and data minimization strategies
- Safeguarding sensitive knowledge sources and PII from unauthorized agent access
Observability, Auditing, and Incident Response
- Building telemetry for agent behavior: intent tracing, command logging, and provenance
- SIEM integration, setting alerting thresholds, and ensuring forensic readiness
- Developing runbooks and playbooks for managing agent-related incidents and containment
Red-Teaming Agentic Systems
- Planning red-team exercises: defining scope, rules of engagement, and safe failover mechanisms
- Adversarial tactics: prompt injection, tool misuse, chain-of-thought manipulation, and API abuse
- Executing controlled attacks and assessing exposure and impact
Hardening and Mitigations
- Engineering safeguards: response throttles, capability gating, and sandboxing
- Policy and orchestration controls: approval workflows, human-in-the-loop mechanisms, and governance hooks
- Model and prompt-level defenses: input validation, canonicalization, and output filtering
Operationalizing Safe Agent Deployments
- Deployment strategies: staging, canary releases, and progressive rollouts for agents
- Change control, testing pipelines, and pre-deployment safety verification
- Cross-functional governance: aligning security, legal, product, and ops playbooks
Capstone: Red-Team / Blue-Team Simulation
- Perform a simulated red-team attack on a sandboxed agent environment
- Defend, detect, and remediate as the blue team using established controls and telemetry
- Present findings, remediation strategies, and policy improvements
Summary and Next Steps
Requirements
- Strong foundation in security engineering, system administration, or cloud operations
- Understanding of AI/ML principles and the behavior of large language models (LLMs)
- Practical experience with identity & access management (IAM) and secure system architecture
Target Audience
- Security engineers and red-team specialists
- AI operations and platform engineers
- Compliance officers and risk management professionals
- Engineering leads overseeing agent deployments
Testimonials (1)
inventory and identifying the different risk exposures within AI
