Helping people use AI safely.

Private AI training, group workshops, and self-paced courses.

Learn with us.

Private and confidential

Private AI training

One to one, with a colleague or two, or for your whole team. At your home, your office or online, in English or Polish.

Choose a single session, a half day, a day with your team or a short series.

  1. You show me how you work. We sit together while you do your real tasks. I observe, ask questions, and plan.
  2. We choose what to fix, adjust or automate.
  3. We build it together, with a focus on safe foundations and playbooks for future use. Sometimes that's a scheduled Codex task or a set of tailored Claude skills. Sometimes we generate a custom automation that does it better and safer, and runs with no AI at all.
  4. You learn AI while working on your own tasks and documents, not abstract examples. You can use it the same day.
  • An NDA is available before we start.
  • Your files and business information remain confidential.

Tell us what you need. We will send a proposal within two working days.

Arrange a private session

Courses and tools.

At your own pace

Self-paced courses

Guides and courses for learning to use Claude safely on your own account, at your own pace. Some are free.

Free tools and other projects

  • Basic ScanA free security scanner: 23 checks on any website.
  • 5D quizCheck your AI fluency in a few minutes.
  • Safe AI on SkoolA community comparing notes on safe AI use.
  • AI for SchoolResources for teachers, parents and school leaders.
  • AEDDAEU-aligned due diligence on AI vendors.
  • Open sourceOur tools and resources, published for inspection.

Case study.

She didn't trust AI. So we started small.

A professional in the marine industry produced weekly and six-monthly market intelligence reports by hand. She gathered news from subscription sources and web searches, translated it into English and formatted it to a fixed structure. The process took hours each week and four to five days every six months.

We created a separate email account and drive for the task, with no sensitive information in either. Then we trained her to build a scheduled Claude task that reviews forwarded emails, selects relevant items, searches the web, formats the report and saves it to her drive. It runs every Sunday night.

She reviews each report, corrects errors and updates the task when her needs change. She has used the system since June.

  • Hours → minutesweekly reports now take
  • 4–5 days → minutessix-monthly reports now take
About the founder

Diana Szyperska

aka Skipper Szyper

Diana came to AI from teaching critical thinking, managing cybersecurity operations, and leading compliance. She now builds AI tools and trains others to use AI safely. Your team gets the technical detail with a focus on safety, explained in their language. AI Safe Lens is independent, self-funded, and not affiliated with any AI company. Any specific tool recommendations come from individual experience, objective evaluations, and clients' preferences.

  • Educator: 20+ years teaching critical thinking, ethics, integrated humanities, and now AI literacy.
  • ISO lead: Implementation lead for ISO 27001 Information Security and ISO 9001 Quality Management Systems.
  • Cyber Security: Operations Manager in a 24/7 Cyber Security Operations Centre (SOC).
Diana Szyperska