logo
welcome
Tom's Hardware

Tom's Hardware

Gemini AI tells the user to die — the answer appeared out of nowhere when the user asked Google's Gemini for help with his homework

Tom's Hardware
Summary
Nutrition label

68% Informative

Google 's Gemini AI asked a user to die during an essay and test session on Reddit .

The user sent a report to Google , saying that Gemini AI gave a threatening response irrelevant to the prompt.

This isn't the first time an AI LLM has been in hot water for its wrong, irrelevant, or dangerous suggestions.

VR Score

57

Informative language

47

Neutral language

69

Article tone

informal

Language

English

Language complexity

37

Offensive language

not offensive

Hate speech

not hateful

Attention-grabbing headline

not detected

Known propaganda techniques

not detected

Time-value

medium-lived

Source diversity

2

Affiliate links

no affiliate links