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: including misuse, privilege escalation, data leakage, and supply-chain risks
- Analyzing adversary profiles and attacker capabilities specific to autonomous agents
- Identifying assets, defining trust boundaries, and pinpointing critical control points for agents
Governance, Policy, and Risk Management
- Implementing governance frameworks for agentic systems, defining roles, responsibilities, and approval gates
- Structuring policies around acceptable use, escalation rules, data handling, and audit capabilities
- Addressing compliance requirements and collecting evidence for audit purposes
Non-Human Identity and Authentication for Agents
- Creating agent identities through service accounts, JWTs, and short-lived credentials
- Applying least-privilege access patterns and just-in-time credentialing mechanisms
- Managing the identity lifecycle, including rotation, delegation, and revocation strategies
Access Controls, Secrets, and Data Protection
- Implementing fine-grained access control models and capability-based patterns for agents
- Overseeing secrets management, encryption during transit and at rest, and data minimization practices
- Safeguarding sensitive knowledge sources and PII from unauthorized agent access
Observability, Auditing, and Incident Response
- Developing telemetry for agent behavior, including intent tracing, command logging, and provenance tracking
- Integrating with SIEM tools, setting alerting thresholds, and ensuring forensic readiness
- Creating 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 procedures
- Utilizing adversarial techniques such as prompt injection, tool misuse, chain-of-thought manipulation, and API abuse
- Executing controlled attacks to assess exposure and impact
Hardening and Mitigation Strategies
- Applying engineering controls like response throttles, capability gating, and sandboxing
- Enforcing policy and orchestration controls, including approval flows, human-in-the-loop checks, and governance hooks
- Implementing model and prompt-level defenses via input validation, canonicalization, and output filtering
Operationalizing Safe Agent Deployments
- Utilizing deployment patterns such as staging, canary releases, and progressive rollouts for agents
- Managing change control, testing pipelines, and pre-deployment safety checks
- Fostering cross-functional governance across security, legal, product, and operations teams
Capstone: Red-Team / Blue-Team Simulation
- Conducting a simulated red-team attack against a sandboxed agent environment
- Acting as the blue team to defend, detect, and remediate using established controls and telemetry
- Presenting findings, outlining a remediation plan, and suggesting policy updates
Conclusion and Future Steps
Requirements
- A robust foundation in security engineering, system administration, or cloud operations
- Understanding of AI/ML concepts and the behavioral characteristics of large language models (LLMs)
- Hands-on experience with identity and access management (IAM) and secure system design
Target Audience
- Security engineers and red-team practitioners
- 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
