# Cybersecurity Courses > Source: https://ibl.ai/courses/cybersecurity > Last updated: 2026-11-14 **Cybersecurity courses from ibl.ai. Securing LLM applications and AI agents: threat modeling, prompt injection, RAG and MCP security, guardrails, PII and model protection, vulnerability disclosure, security testing, and the governance frameworks (EU AI Act, NIST AI RMF, ISO/IEC 42001).** Subjects: Information Technology · Computer Science. Other catalogs: [Higher Education](https://ibl.ai/courses/higher-education). ## Courses ### [AI Governance, Compliance and Audit Trails: EU AI Act, NIST AI RMF and ISO/IEC 42001](https://ibl.ai/courses/cybersecurity/ai-compliance-and-audit-trails) A hands-on, self-paced course for security, platform and compliance teams on AI governance evidence: the EU AI Act as amended in July 2026, the NIST AI RMF and Generative AI Profile, ISO/IEC 42001 and SOC 2 at the level their publishers describe, and five local labs that build and test a tamper-evident audit log for an AI agent. Free · Self-paced, online · 10 to 12 hours · 3 hours per week · Intermediate · English · Information Technology ### [Continuous Security Testing for LLM Apps: Automated Evaluation and CI Gates](https://ibl.ai/courses/cybersecurity/continuous-security-testing-for-llm-apps) A hands-on, self-paced course for security, AppSec, platform and ML teams on making security testing of LLM applications continuous: authorized scope and rules of engagement, framework scans with garak and PyRIT, a harmless canary and policy regression suite, thresholds and trial counts, and a CI gate that blocks a merge when the failure rate is too high, taught from OWASP, NIST and MITRE ATLAS. Free · Self-paced, online · 10 to 12 hours · 3 hours per week · Intermediate · English · Information Technology ### [Coordinated Vulnerability Disclosure for AI Systems](https://ibl.ai/courses/cybersecurity/ai-vulnerability-disclosure) A hands-on, self-paced course for security, product security and ML teams on running a vulnerability disclosure program for AI products: a policy with safe harbor language, security.txt, intake that separates security from model behavior, triage and severity, canary regression tests, coordination and publication through CVE and CSAF, taught from CISA, CVE, OASIS, OWASP, MITRE ATLAS and NIST sources. Free · Self-paced, online · 10 to 12 hours · 3 hours per week · Intermediate · English · Information Technology ### [Defending Against Prompt Injection: Direct, Indirect and Multi-Turn](https://ibl.ai/courses/cybersecurity/defending-against-prompt-injection) A hands-on, self-paced course for security and engineering teams on defending LLM applications and agents against direct, indirect and multi-turn prompt injection. Measure exposure with harmless canaries on a local sample app, add least-privilege tools, human approval, output handling and structured input one at a time, and build a conversation-level monitor and regression suite, taught from OWASP, NIST and MITRE ATLAS. Free · Self-paced, online · 10 to 12 hours · 3 hours per week · Intermediate · English · Information Technology ### [FBI FLASH: Defending Against ARTEX Autonomous Intrusions](https://ibl.ai/courses/cybersecurity/fbi-flash-artex-defense) A hands-on, self-paced course built from the FBI FLASH-20261113-001 (13 November 2026, TLP:CLEAR) on the ARTEX autonomous penetration testing framework. Map the FBI's observed activity to MITRE ATT&CK, inventory peripheral applications, restrict administration and publishing paths, review service principals and consent grants, build detections over synthetic logs, vet the FBI's indicators before acting, and prepare an evidence-first response and FBI report pack. Free · Self-paced, online · 8 to 10 hours · 3 hours per week · Intermediate · English · Information Technology ### [Identity and Delegated Access for AI Agents: OAuth, Scoped Tokens and Human Approval](https://ibl.ai/courses/cybersecurity/identity-and-delegated-access-for-agents) A hands-on, self-paced course for security, identity and platform teams on giving AI agents access safely: why agents must never hold users' passwords or standing admin keys, OAuth 2.0 under RFC 9700, token exchange, scoped and short-lived tokens, the MCP authorization specification, human approval and revocation, with labs on a local authorization broker. Free · Self-paced, online · 10 to 12 hours · 3 hours per week · Intermediate · English · Information Technology ### [Interpretability for AI Assurance: Inspecting Model Internals](https://ibl.ai/courses/cybersecurity/interpretability-for-ai-assurance) A hands-on, self-paced course for security and ML teams on mechanistic interpretability as assurance evidence. Learn why black-box testing leaves gaps, run activation caching, the logit lens, activation patching and probing on GPT-2 small with TransformerLens, read what research found on sycophancy, reward hacking and hidden objectives, and write an assurance plan that states what interpretability can and cannot certify. Free · Self-paced, online · 10 to 12 hours · 3 hours per week · Intermediate · English · Computer Science ### [Output Validation and Guardrails: Structured Outputs, Constrained Decoding and Sanitization](https://ibl.ai/courses/cybersecurity/output-validation-and-guardrails) A hands-on, self-paced course for security and engineering teams on controlling what LLM applications output: OWASP LLM05 and LLM09, structured outputs and constrained decoding, local schema validation, sanitizing for browsers, shells, databases and tools, guards that fail closed, NeMo Guardrails, and logging and testing every decision, in eight local labs on synthetic data. Free · Self-paced, online · 10 to 12 hours · 3 hours per week · Intermediate · English · Information Technology ### [Protecting Models and Training Data: Extraction, Membership Inference and Defenses](https://ibl.ai/courses/cybersecurity/protecting-models-and-training-data) A hands-on, self-paced course for security, platform and ML teams that serve models through an API: what extraction, membership inference and weights theft achieve and the evidence they leave, then gateway quotas, detection of systematic querying, differential privacy in training and signed artefact checks, taught from OWASP, NIST and MITRE ATLAS. Free · Self-paced, online · 10 to 12 hours · 3 hours per week · Intermediate · English · Information Technology ### [Protecting PII in LLM Applications: Masking, Tokenization and Privacy Proxies](https://ibl.ai/courses/cybersecurity/protecting-pii-in-llm-applications) A hands-on, self-paced course for security, AppSec, platform and ML teams on protecting personal data in LLM applications: map the data flows, minimise, measure PII detection on a synthetic test set, build a privacy proxy that tokenizes values before a model sees them and restores them after, and log without leaking, taught from OWASP, NIST, MITRE ATLAS and primary legal texts. Free · Self-paced, online · 10 to 12 hours · 3 hours per week · Intermediate · English · Information Technology ### [Securing Agent Tools and MCP Servers](https://ibl.ai/courses/cybersecurity/securing-mcp-servers-and-agent-tools) A hands-on, self-paced course for security, platform and ML teams on securing MCP servers and agent tools: tool poisoning, look-alike tools, rug pulls, over-broad scopes and the confused deputy, and the server-side controls that stop them. Build and test a hardened server with the official Python SDK, then write a tool security review and hardening plan. Free · Self-paced, online · 10 to 12 hours · 3 hours per week · Intermediate · English · Information Technology ### [Securing RAG: Access Control, Provenance and Poisoning Defenses](https://ibl.ai/courses/cybersecurity/securing-rag-pipelines) A hands-on, self-paced course for security, platform and ML teams on securing retrieval augmented generation: permission-aware retrieval, ingestion controls, signed provenance, citation verification, and detecting poisoned documents with harmless canaries, taught from OWASP LLM08 and LLM04, NIST, MITRE ATLAS, CISA and the vector databases' own documentation. Free · Self-paced, online · 10 to 12 hours · 3 hours per week · Intermediate · English · Information Technology ### [Threat Modeling AI Agents and LLM Applications](https://ibl.ai/courses/cybersecurity/threat-modeling-ai-agents) A hands-on, self-paced course for security engineers, AppSec, platform and ML teams on threat modeling LLM applications and tool-using agents: data-flow diagrams and trust boundaries, STRIDE with the OWASP catalogues and MITRE ATLAS, the confused deputy, and a control, test and monitoring signal for every threat. Free · Self-paced, online · 10 to 12 hours · 3 hours per week · Intermediate · English · Information Technology [Browse on lms.ibl.ai](https://lms.ibl.ai/platform/cybersecurity/discover)