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
- Categorizing agentic threats: misuse, privilege escalation, data leakage, and supply-chain vulnerabilities
- Defining adversary profiles and attacker capabilities unique to autonomous agents
- Mapping assets, trust boundaries, and critical control points for agent interactions
Governance, Policy Frameworks, and Risk Management
- Implementing governance frameworks for agentic systems, including roles, responsibilities, and approval gates
- Designing policies covering acceptable use, escalation protocols, data handling, and auditability
- Addressing compliance requirements and gathering evidence for audits
Non-Human Identity & Authentication for Agents
- Structuring agent identities using service accounts, JWTs, and short-lived credentials
- Applying least-privilege access patterns and just-in-time credentialing strategies
- Managing identity lifecycle processes, including rotation, delegation, and revocation
Access Controls, Secret Management, and Data Protection
- Utilizing fine-grained access control models and capability-based patterns for agents
- Implementing secrets management, encryption (in-transit and at-rest), and data minimization practices
- Safeguarding sensitive knowledge sources and PII from unauthorized agent access
Observability, Auditing, and Incident Response
- Engineering telemetry for agent behavior, including intent tracing, command logging, and provenance tracking
- Integrating with SIEM tools, setting alerting thresholds, and maintaining forensic readiness
- Developing runbooks and playbooks for handling agent-related incidents and containment
Red-Teaming Agentic Systems
- Planning red-team engagements: defining scope, rules of engagement, and safe failover mechanisms
- Applying adversarial techniques such as prompt injection, tool misuse, chain-of-thought manipulation, and API abuse
- Executing controlled attacks to measure system exposure and impact
Hardening and Mitigation Strategies
- Deploying engineering controls like response throttles, capability gating, and sandboxing
- Enforcing policy and orchestration controls through approval flows, human-in-the-loop mechanisms, and governance hooks
- Implementing model and prompt-level defenses via input validation, canonicalization, and output filtering
Operationalizing Safe Agent Deployments
- Adopting deployment patterns such as staging, canary releases, and progressive rollouts for agents
- Managing change control, testing pipelines, and pre-deployment safety checks
- Facilitating cross-functional governance through coordinated security, legal, product, and ops playbooks
Capstone Project: Red-Team vs. Blue-Team Exercise
- Simulating a red-team attack against a sandboxed agent environment
- Performing defense, detection, and remediation as the blue team using established controls and telemetry
- Presenting findings, remediation plans, and necessary policy updates
Conclusion and Future Directions
Requirements
- Strong foundation in security engineering, system administration, or cloud operations
- Proficiency with AI/ML concepts and the behavioral patterns of large language models (LLMs)
- Hands-on 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
