Appearance
← All courses

ISO/IEC 27090 · Independent introductory draft

AI security threats and safeguards

Understand AI attack surfaces, poisoning, evasion, extraction, privacy exposure and lifecycle safeguards.

Start with the syllabus chapters

Study 8 dedicated topic chapters before using the domain overviews and practice below.

Open the ISO/IEC 27090 syllabus chapters →

Foundation modules and applied practice

Read the lessons in order, try each fictional scenario, then use the practice questions to identify concepts to revisit. Reading and practice are free. No academy account is required.

Your ISO/IEC 27090 progress

Loading progress…

Saving is off; progress stays in this course session only. No academy account or cloud sync. Each course has separate data.

Individual quiz answers remain in memory only. Completed score summaries can be saved and backed up. Storage details

Practice history

No completed attempts in this course history.

8 matching lessons

  1. Academy module 1 · ISO/IEC 27090 · About 8 minutes

    Map AI assets and lifecycle trust boundaries

    Track training evaluation and operating inputs.

    Not marked complete · Not bookmarked

  2. Academy module 2 · ISO/IEC 27090 · About 8 minutes

    Understand poisoning and backdoor risk

    Understand where data and labels came from.

    Not marked complete · Not bookmarked

  3. Academy module 3 · ISO/IEC 27090 · About 8 minutes

    Evaluate evasion and robustness carefully

    Define attacker access and permitted input changes.

    Not marked complete · Not bookmarked

  4. Academy module 4 · ISO/IEC 27090 · About 8 minutes

    Protect model artifacts and inference access

    Restrict model stores signing and deployment identities.

    Not marked complete · Not bookmarked

  5. Academy module 5 · ISO/IEC 27090 · About 8 minutes

    Distinguish AI privacy attack objectives

    Infer whether a record participated in training.

    Not marked complete · Not bookmarked

  6. Academy module 6 · ISO/IEC 27090 · About 8 minutes

    Apply trust boundaries to generative AI tools

    Treat retrieved text and user input as data.

    Not marked complete · Not bookmarked

  7. Academy module 7 · ISO/IEC 27090 · About 8 minutes

    Control model supply chains and updates

    Record component data and model origins.

    Not marked complete · Not bookmarked

  8. Academy module 8 · ISO/IEC 27090 · About 8 minutes

    Investigate AI-specific security signals

    Collect relevant protected observations.

    Not marked complete · Not bookmarked

Practice with explanations

16 original questions, with two questions per foundation module. Every option has an explanation and a related lesson link. These are not copied exam questions.

Start ISO/IEC 27090 practice →

Active answers remain in memory while navigating within this course. Refreshing or leaving the course resets the attempt. Completed summaries can be saved in this course’s separate history and backups.

Practice with ISO/IEC 27090 flashcards · Applied frameworks and Indian regulatory context

Edition and applied coverage map

Further study still needed

Official orientation and background reading

Lessons, scenarios and questions are original Safehouse learning material. Source links provide orientation and background; they do not certify completeness.

Found an issue? Report a correction.