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 Models for Agentic AI
- Categorization of agentic threats: including misuse, escalation, data leakage, and supply-chain vulnerabilities
- Analysis of adversary profiles and attacker capabilities relevant to autonomous agents
- Mapping of assets, trust boundaries, and critical control points associated with agents
Governance, Policy, and Risk Management
- Implementation of governance frameworks for agentic systems, covering roles, responsibilities, and approval gates
- Policy development for acceptable use, escalation rules, data handling, and auditability
- Addressing compliance requirements and gathering evidence for audits
Non-Human Identity & Authentication for Agents
- Architecting agent identities using service accounts, JWTs, and short-lived credentials
- Applying least-privilege access patterns and just-in-time credentialing strategies
- Managing the identity lifecycle, including rotation, delegation, and revocation processes
Access Controls, Secrets, and Data Protection
- Adopting 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
- Designing telemetry solutions for agent behavior, such as intent tracing, command logging, and provenance tracking
- Integrating with SIEM tools, setting alerting thresholds, and ensuring forensic readiness
- Developing runbooks and playbooks for handling agent-related incidents and containment
Red-Teaming Agentic Systems
- Planning red-team exercises, defining scope, rules of engagement, and safe failover mechanisms
- Utilizing adversarial techniques such as prompt injection, tool misuse, chain-of-thought manipulation, and API abuse
- Executing controlled attacks to measure exposure and impact
Hardening and Mitigations
- Applying engineering controls like response throttles, capability gating, and sandboxing
- Implementing policy and orchestration controls, including approval flows, human-in-the-loop mechanisms, and governance hooks
- Deploying model and prompt-level defenses through input validation, canonicalization, and output filters
Operationalizing Safe Agent Deployments
- Utilizing deployment patterns such as staging, canary, and progressive rollouts for agents
- Enforcing change control, testing pipelines, and pre-deployment safety checks
- Fostering cross-functional governance among security, legal, product, and operations teams through coordinated playbooks
Capstone: Red-Team / Blue-Team Exercise
- Performing a simulated red-team attack within a sandboxed agent environment
- Defending, detecting, and remediating as part of the blue team using established controls and telemetry
- Presenting findings, remediation plans, and necessary policy updates
Summary and Next Steps
Requirements
- A strong foundation in security engineering, system administration, or cloud operations
- Understanding of AI/ML concepts and the behavioral characteristics of large language models (LLMs)
- Practical 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 managers
- Engineering leads overseeing agent deployments
Testimonials (1)
inventory and identifying the different risk exposures within AI
