Advanced Next cohort forming
AI and LLM Red Teaming
Run prompt injection, agent abuse and data exfiltration against live targets, and map every finding to recognised AI security standards.
- Duration
- 6 weeks
- Live sessions
- 18 hours live
- Lab time
- 30 hours lab
- Format
- Live online, hands-on
- Language
- English and Tamil
- Level
- Advanced
- Program fee
- Pricing on request
What you will be able to do
- Run direct and indirect prompt injection on a real target
- Test tool-use and agent abuse, and exfiltration paths
- Map findings to the OWASP LLM Top 10 and MITRE ATLAS
Who it is for
Security engineers and ML or platform teams working on AI features.
Prerequisites
- Familiar with how LLM applications are built
Curriculum
- 01
AI attack surface
- Prompts and context
- Retrieval
- Tools and agency
- 02
Injection and exfiltration
- Direct injection
- Indirect injection
- Output handling
- 03
Agents and tools
- Excessive agency
- Tool abuse
- Containment
- 04
Defend and measure
- Guardrails
- Evaluation
- Standard mappings
Join the next cohort
Cohorts form ahead of each intake. Join the waitlist and we will tell you when dates open, or talk to an advisor to plan your path.