Protecting models, data, applications, agents and AI infrastructure: threat modeling, prompt injection, adversarial ML, red teaming, supply chain and privacy.
Use the system you need to protect to select your learning focus. Public topics address workplace data handling, prompt injection, agent tools, model attacks, supply chains and infrastructure. Compare introductory safety material with deeper engineering courses according to your role. Read the syllabus for the assets and threats covered rather than assuming one course addresses every risk. Individual pages show availability and prerequisites so you can choose an appropriate security study path.
on sale today, with every lesson, exam and certificate. If you are signed in, courses you already own are left out and the price at checkout drops to match.
Only the courses on sale in this track today are included. Courses added to the track later are not part of the bundle; buy them separately, or get the All-Access Pass, which includes new courses while it lasts.
Follow one fictional US distributor through four everyday AI mistakes (a guarantor's credit file uploaded to a personal account, a customer letter promising a fee waiver that does not exist, a vendor email with hidden instructions for the assistant, and a note-taker that sent a customer the internal margin floor) and practice the decisions that prevent or contain each one in three short Decision Labs.
A plain-language course on handling data safely with AI tools: what happens to what you type or upload, how to classify before you paste, how connected apps and agents change the picture, and how to report a mistake quickly.
6 h · 15 lessonsSee price at checkout
Intermediate
For practitioners who use or build with AI. · 5 courses
A practitioner course on securing applications that use large language models, organized around the 2026 OWASP Top 10 for LLM Applications. Every risk comes with a hypothetical scenario, the root cause, and concrete defenses in code and configuration.
A deep course on prompt injection: why it exists, how direct, indirect and multimodal payloads turn into data exfiltration and goal hijack in agents, why filters alone fail against adaptive attackers, and which architectural patterns and tests actually reduce risk.
A hands-on course in threat modeling chat, RAG and agent features: data flow diagrams with real trust boundaries, threat enumeration with STRIDE, MITRE ATLAS and NIST AI 100-2, risk rating for probabilistic components, and secure design patterns that hold up when the model is fooled.
A practical course in agent security: the OWASP Top 10 for Agentic Applications, identity and delegated authorization for agents, securing MCP clients and servers under the 2026-07-28 specification, sandboxed execution, multi-agent communication, and the logging and kill switches that make agents operable.
A practitioner course on AI data security across the lifecycle: provenance and integrity for training data, poisoning prevention, permission-aware retrieval and multi-tenant vector store isolation, leakage through prompts, outputs and logs, and realistic vendor retention and deletion controls.
An expert course on authorized AI red teaming: scoping and rules of engagement, threat-informed planning with MITRE ATLAS, attacking LLM applications, RAG and agents, automating multi-turn attacks at scale, measuring results with statistical care, and running a continuous program.
An expert course on AI supply chain security: the threats to base models, adapters, datasets, ML libraries, containers and agent tools; safe serialization and model scanning; signing with the OpenSSF Model Signing specification and Sigstore; SLSA provenance; AI bills of materials; and a registry with promotion gates.
An expert course on privacy-preserving machine learning: privacy attacks on models, differential privacy and DP-SGD, federated learning and secure aggregation, homomorphic encryption, secure computation and confidential computing, privacy in LLM applications, and how to audit privacy claims.
An expert course on adversarial machine learning organized around the NIST AI 100-2 E2025 taxonomy: evasion attacks from FGSM and PGD to black-box and physical-world attacks, poisoning and backdoors, model extraction from prediction and LLM APIs, and defenses evaluated against adaptive attackers.
An expert course on securing the infrastructure underneath AI: GPU cluster multi-tenancy and scheduler hardening, fabric and storage access control, inference server and API hardening, protecting model weights, MLOps pipeline and registry security, GPU confidential computing, and control mapping to NIST and CIS.
An expert course on defending AI systems in production: mapping ATLAS techniques to observable telemetry, logging that supports forensics without creating new privacy risk, detections for injection, agent misbehavior and model abuse, incident response playbooks, forensics under non-determinism, and recovery.