Building Secure and Responsible LLM Applications Training Course

LLM Architecture and Attack Vectors

• Construction, deployment, and API access methods for LLMs.
• Core components of LLM application stacks (prompts, agents, memory, APIs).
• Identifying where and how security issues manifest in real-world scenarios.

Prompt Injection and Jailbreak Vulnerabilities

• Defining prompt injection and assessing its potential danger.
• Analyzing direct and indirect prompt injection scenarios.
• Techniques used for jailbreaking to bypass safety filters.
• Strategies for detecting and mitigating these attacks.

Data Leakage and Privacy Threats

• Preventing accidental data exposure in model responses.
• Addressing PII leaks and improper use of model memory.
• Designing privacy-focused prompts and Retrieval-Augmented Generation (RAG) systems.

LLM Output Filtering and Safeguards

• Applying Guardrails AI for content filtering and validation.
• Establishing output schemas and constraints.
• Monitoring and logging for unsafe outputs.

Human-in-the-Loop and Process Integration

• Determining optimal points for introducing human oversight.
• Managing approval queues, scoring thresholds, and fallback procedures.
• Calibrating trust and leveraging explainability.

Secure LLM Application Design Patterns

• Implementing least privilege and sandboxing for API calls and agents.
• Applying rate limiting, throttling, and abuse detection mechanisms.
• Building robust chains with LangChain and ensuring prompt isolation.

Compliance, Logging, and Governance

• Ensuring the auditability of LLM outputs.
• Maintaining traceability through prompt and version control.
• Aligning with internal security policies and regulatory requirements.

Conclusion and Future Directions