Skip to main content

Topic roadmap

Security

32 lessons, about 97.5 hours in all, from the basics to advanced. Start with the first, or jump to what you need.

Lessons, easiest first

  1. Security · Beginner 1 · about 3 hUse AI safely at workDecide what you may and may not paste into an AI tool at work, and explain why.
  2. Security · Beginner 2 · about 3 hHow cyber attacks workExplain the CIA triad, who attacks and why, what an attack surface is, the difference between a vulnerability, an exploit and a CVE, and the stages of an attack from reconnaissance to impact.
  3. Security · Beginner 2 · about 3 hLinux for securityRead Linux users, groups and file permissions, find running processes and services, use sudo and SSH keys safely, and search the logs in /var/log for signs of an attack.
  4. Security · Beginner 2 · about 3 hNetworking for securityExplain IP addresses, ports, TCP and UDP, DNS, HTTP and HTTPS and firewalls well enough to read a connection log and say what looks normal and what looks suspicious.
  5. Security · Beginner 3 · about 3 hCryptography basicsTell hashing from encryption, explain symmetric and asymmetric keys, describe what TLS and certificates protect, store passwords with a slow hash such as Argon2 or bcrypt, and keep secrets out of code, logs and prompts.
  6. Security · Beginner 3 · about 3 hIdentity and accessTell authentication from authorisation, choose sensible password and MFA rules, explain OAuth and OpenID Connect at a high level, apply least privilege with IAM roles, and keep sessions secure.
  7. Security · Beginner 3 · about 4 hWeb app security and the OWASP Top 10Explain the OWASP Top 10 web risks, recognise SQL injection, cross-site scripting and broken access control in code, and fix each one with the standard defence.
  8. Security · Intermediate 1 · about 3 hIncident response basicsRun the first hours of a security incident: confirm and scope it, contain it, preserve evidence, meet India's CERT-In reporting duty, recover and learn from it.
  9. Security · Intermediate 1 · about 3 hLogs and SIEMDecide what a system should log, bring logs from many sources into a SIEM, and turn them into a small number of alerts an analyst can act on.
  10. Security · Intermediate 1 · about 3 hPhishing and social engineeringRecognise phishing, spear phishing, business email compromise and vishing, set up SPF, DKIM and DMARC, and design a reporting and response flow that works even when AI makes the lure convincing.
  11. Security · Intermediate 1 · about 3 hThreat detection with MITRE ATT&CKMap attacker behaviour to MITRE ATT&CK, write and tune a simple detection rule in Sigma, and run a basic hypothesis-driven threat hunt.
  12. Security · Intermediate 2 · about 2.5 hAI-powered attacks and how defenders respondExplain how attackers use AI for phishing, voice and video impersonation, reconnaissance, malware and finding vulnerabilities, and put in place the process and technical controls that still work when the fake is convincing.
  13. Security · Intermediate 2 · about 3 hAudit logsRecord who asked what, which data the AI system touched, what it answered or did and why, in a tamper-resistant log that supports investigations and audits without leaking personal data.
  14. Security · Intermediate 2 · about 3 hCloud security basicsExplain the shared responsibility model, find and fix the common cloud misconfigurations (public storage, over-broad IAM, exposed keys, open ports), turn on cloud logging and use a CSPM tool to keep an account hardened.
  15. Security · Intermediate 2 · about 3 hLLMs in the SOCUse an LLM copilot to summarise alerts, explain logs, draft detection queries and incident reports, while guarding against made-up indicators, leaked logs and prompt injection, and measure the time it really saves.
  16. Security · Intermediate 2 · about 3 hMachine learning for threat detectionExplain how anomaly detection and classifiers find suspicious logins, traffic, phishing and malware, size their false positives with the base rate, and tune a detector a SOC can live with.
  17. Security · Intermediate 2 · about 4 hSingle sign-onPlug an AI app into a company's login with OpenID Connect or SAML, read the user's identity and groups from tokens, and pass that identity safely to downstream systems.
  18. Security · Intermediate 3 · about 3 hThreats to LLM apps: the OWASP Top 10Walk any LLM app through the ten OWASP risks for LLM applications, name a real example and the main fix for each, and know which lesson goes deeper on each risk.
  19. Security · Pro 1 · about 3 hData poisoning, backdoors and adversarial examplesExplain how attackers corrupt a model through its training data, fine-tuning data or knowledge base, or fool it with crafted inputs, and put controls at each step of the pipeline.
  20. Security · Pro 1 · about 3 hDefend against prompt injectionExplain prompt injection, direct and indirect, and design an app so an injection cannot do real harm.
  21. Security · Pro 1 · about 3 hGuardrailsAdd input and output checks around a model: topic limits, personal data masking, format validation and policy checks, choosing rules or models for each.
  22. Security · Pro 1 · about 3 hIncident response for AIDetect, contain, fix and learn from AI failures in production, from outages and cost spikes to harmful or wrong outputs, with roles, runbooks and blameless reviews.
  23. Security · Pro 1 · about 3 hModel extraction and privacy attacksExplain how attackers copy a model or pull private training data out of it through its API, and design the API, training data and monitoring so they cannot.
  24. Security · Pro 1 · about 3 hPermissions and sandboxingGive an agent the least access it needs, act with the user's own permissions, and run any code it writes in an isolated sandbox.
  25. Security · Pro 1 · about 3 hSecuring agent tools and MCP serversSpot confused-deputy and tool-poisoning risks in an agent's tools and MCP servers, and add allow-lists, scoped credentials, approval screens and logging so a bad tool or tool result cannot cause real harm.
  26. Security · Pro 1 · about 3 hSecuring the AI supply chainCheck models, datasets and packages before you use them, load model files safely, pin what you depend on, and keep an AI bill of materials that says exactly what is in your system.
  27. Security · Pro 1 · about 2 hSecurity questionnairesAnswer a client's vendor security questionnaire accurately and quickly, using a maintained answer library, evidence and a clear description of your AI system's data flows.
  28. Security · Pro 1 · about 3 hThreat model an AI systemDraw the data flow of an AI system, find threats to each part with STRIDE and MITRE ATLAS, rank them, and turn the top ones into controls and tests.
  29. Security · Pro 2 · about 4 hRed-team an AI systemPlan and run an authorised red-team exercise on an AI app: scope, rules of engagement, manual and automated attacks with PyRIT, garak and promptfoo, attack success rate, findings and retests.
  30. Security · Pro 2 · about 3 hRun an AI security programmeSet up AI security across a company: an AI inventory, handling shadow AI, an acceptable-use policy, vendor and feature reviews, metrics, training and clear owners.
  31. Security · Pro 2 · about 3 hSecurity and privacyProtect API keys, user data and access in an LLM app, using the OWASP list as a checklist.
  32. Security · Pro 2 · about 3 hSecurity tests in CI for AI appsRun AI security tests on every change: a fixed attack suite with pass thresholds, regression tests for every fixed finding, and scans of dependencies, secrets and model files, wired into GitHub Actions.

On these role roadmaps

Interview questions on security

Books on security

All recommended books

Today's AI turns are used

Lessons, roadmaps, videos, practice cards and the code playground stay free with your account, with no limit.