About this course
Explore security risks introduced by LLM applications and autonomous agents in this free IBM video course. The selected explainers examine prompt injection, excessive privileges, exposed data and delegation. The emphasis is recognizing trust boundaries and designing safeguards, rather than treating one product or framework as a complete security solution.
What you can learn
- Prompt injection risk
- Agent identity
- Least privilege
- Zero-trust concepts
- AI data protection
- Agent threat modeling
Course at a glance
| Provider | IBM |
|---|---|
| Platform | YouTube · IBM Technology |
| Access | Free Video |
| Level | Intermediate — Coursya audience assessment |
| Language | English |
| Selected lessons | 23 |
| Video duration | 5h 5m · allow extra time for reflection and practice |
| Details checked | 9 October 2026 |
Inside the course
Review agent and LLM vulnerabilities, then connect tokens, delegation and identity to secure architecture. Later lessons discuss hidden instructions, shadow AI, model-service abuse and the relationship between security and governance.
Try this learning exercise
Map a document-reading assistant that can call an external tool. Mark untrusted inputs, sensitive data and privileged actions, then propose approval, isolation and monitoring controls for each boundary.
How to follow the course
Choose a lesson from the list or search for a specific topic. Use Previous and Next to continue, and mark lessons complete as you go. Progress is saved in this browser without a Coursya account.
Each course retains its relevant lessons, including videos also used in another course. Numbered series follow their original lesson or part order. Only reviewed advertisements, unrelated items and identical repeated playlist entries are excluded. Source recordings may have been made at different times.
Unavailable recordings are listed in Course content. Known lesson titles are retained; when the source hides a title or position, this is stated explicitly. Unavailable videos are not included in the video duration.
Free access, requirements and certificates
The selected public YouTube videos are free to watch. No paid subscription, payment card or examination is required for this video learning path. Coursya independently organizes IBM Technology explainers and tutorials for self-paced study. This listing is not an official IBM qualification and does not include an IBM certificate, graded assignment or instructor feedback.
Knowledge of LLM applications and basic cybersecurity concepts is recommended. Examples are for defensive analysis; no paid tool or live attack is required.
Official IBM Technology playlist source ↗
Related courses for your next step
Provider: IBM. Videos play from the IBM Technology channel on YouTube.
Report a correction
Questions & discussion
Share a useful question or correction. Comments appear after moderation. Please avoid personal or sensitive information.