↑

The cybersecurity & AI learning hub.

Free, role-based starting points for everyone from kids and career changers to analysts, forensic investigators, and executives, plus plain-English guides to passwords and building your own artificial intelligence (AI) testing lab.

Learning Paths

Pick your role, and we will point you to the best place to start.

Every path favors free resources. Each one links to the full Resource Library filtered for that role, so you can go deeper when you are ready.

AI & Machine Learning Careers

Free learning for every AI and machine learning role.

AI teams need more than data scientists. These free, vendor-neutral resources cover the technical, infrastructure, application, security, and leadership sides of AI. See the full AI skills list →

Stay Current

Ten minutes a week to stay informed.

Follow the U.S. Cybersecurity and Infrastructure Security Agency (CISA) on LinkedIn or X for alerts, read the Federal Trade Commission business blog for scam warnings, and add one podcast such as Darknet Diaries. Then follow Nexus Security Advisors on LinkedIn for plain-English takeaways. See all recommended news, podcasts, and social accounts →

Everyone

Passwords, passphrases, multi-factor authentication, and encryption in plain English.

Most break-ins start with a stolen or guessed password. These four habits stop the majority of them, and none require technical skills.

1. Length beats complexity

A passphrase of four or more random words, such as orbit-violin-pepper-harbor, is longer, easier to remember, and much harder to crack than P@ssw0rd1!. Aim for 15 characters or more on important accounts. Never reuse a password, and change it when there is a sign it was exposed, not on a fixed calendar. This matches current federal guidance from the National Institute of Standards and Technology (NIST).

2. Let a password manager remember for you

A password manager creates and stores a unique, strong password for every account, so you only remember one master passphrase. Businesses should use a team edition with shared vaults instead of spreadsheets or sticky notes. Popular options include Bitwarden and KeePassXC.

3. Turn on multi-factor authentication (MFA)

Multi-factor authentication, sometimes called two-factor authentication (2FA), asks for a second proof that it is really you. From strongest to weakest:

  • Best: passkeys or physical security keys, which cannot be phished
  • Good: an authenticator app that asks you to match a number
  • Better than nothing: a code sent by text message

Never approve a login request you did not start.

4. Encrypt what matters

  • Devices: turn on built-in disk encryption (BitLocker on Windows, FileVault on Mac, and the default on modern phones) so a lost laptop is not a data breach.
  • Connections: only enter information on sites showing the padlock (HTTPS).
  • Backups: encrypt backups and keep one copy offline or unchangeable.
  • Sensitive files and email: use your email and file platform's built-in encryption or confidential labels.

Online checkers such as Security.org and How Secure Is My Password? show how quickly a password can be cracked. Test a similar example, never your real password.

Hands-On

Set up your own AI testing lab in an afternoon.

A private lab lets you learn how AI models behave, and how they can be attacked, without sending company data to an outside service. Everything below is free.

1. Pick a machine

A modern laptop with 16 GB of memory runs small models. A graphics card, or a short-term cloud server with one, handles larger models. Keep the lab separate from company data and networks.

2. Install a local model runner

Install Ollama or LM Studio, then download a small open model from the Hugging Face hub or the runner's library.

3. Add a chat interface

Use Open WebUI for a familiar chat screen, and notebooks in Jupyter for experiments.

4. Practice attacks safely

Try prompt-injection challenges in Gandalf and PortSwigger's labs, then scan your local model with garak or PyRIT.

5. Add defenses and compare

Put guardrails such as NeMo Guardrails or large language model (LLM) Guard in front of the model and rerun your tests. Map findings to the OWASP Top 10 for AI applications.

Only test systems you own or have written permission to test.

Mentoring & Training

Want a guide for your team?

Nexus Security Advisors offers leadership workshops, tabletop exercises, and role-based security and AI training for small and mid-sized businesses, and mentors veterans and newcomers entering the cybersecurity field.

Let's Talk

Start with a free 30-minute discovery call.

Tell us your industry, size, and the security or compliance challenge in front of you. We'll tell you plainly whether we're a fit.