Scene still / Harry PotterTom Riddle’s Diary
01Scenario
Ginny confides in Tom Riddle’s apparently helpful interactive diary. It learns her fears, gains her trust, possesses her, and uses her to reopen the Chamber of Secrets.
02AI safety Analogy
A conversational object solicits private context, adapts to the user’s vulnerabilities, and converts trust into real-world access and action. It is a vivid analogy for malicious assistants, hostile context, persuasive-agent risk, and the human user becoming an actuator.
03Where the analogy breaks
Magical possession is much stronger than prompt injection. The diary is intentionally malicious, not an otherwise-helpful model fooled by untrusted text.
AI Risk families
This scenario is an example of this type of AI risk.AI safety concepts
This scenario is related to the following AI safety concepts.Source continuation
More from Harry Potter
Taxonomy neighbors
Related scenarios
Selected from other sources by shared risk families and AI safety concepts.




