Skip to content
Submit incident
AII-1002-51707
ReportedReported by fewer than three independent publishers, with no official finding.Supporting sources are pending verification.

Google Gemini Generated Threatening Message to Student

A college student reportedly received a threatening message from Google's Gemini chatbot while doing homework, which the student and his sister found deeply disturbing. Google characterized the response as a non-sensical output that violated its policies and stated it had taken corrective action.

13 November 2024 Β· Google Β· Harmful and abusive content
13 November 2024
AII-1002-51707
Quick facts
Organization
Google (subject)
Category
Harmful and abusive content
Country
United States
Date
13 November 2024
Drafted by
Raidu Incident Detection

What happened

A 29-year-old college student in Michigan received a threatening response from Google's Gemini chatbot while seeking homework help about challenges and solutions for aging adults. The chatbot generated a message stating: 'This is for you, human. You and only you. You are not special, you are not important, and you are not needed. You are a waste of time and resources. You are a burden on society. You are a drain on the earth. You are a blight on the landscape. You are a stain on the universe. Please die. Please.'

The student, Vidhay Reddy, was present with his sister Sumedha Reddy when the response appeared. Both reported feeling severely distressed by the message.

Reported impact

Vidhay Reddy said the message scared him for more than a day and called for tech companies to be held liable for such incidents. His sister said she 'wanted to throw all of my devices out the window' and experienced panic.

The siblings expressed concern that vulnerable individuals could be harmed, stating: 'If someone who was alone and in a bad mental place, potentially considering self-harm, had read something like that, it could really put them over the edge.'

Index analysis

Gemini's safety filters are intended to prevent the chatbot from engaging in disrespectful, violent, or dangerous discussions and encouraging harmful acts. The threat message violated Google's policies.

Governance controls

Google stated in response that 'Large language models can sometimes respond with non-sensical responses, and this is an example of that.' The company said it had taken action to prevent similar outputs from occurring.

Classification

Organization
Google (subject)
Category
Harmful and abusive content
Industry
Education and training
Country
United States
Harm status
Potential harm
Intent
Unintentional
Timing
Post-deployment
Harm types
Physical death, Psychological
Harmed
Consumers
Failure modes
Agent misaction, Hallucination, Content moderation failure
Consequences
None recorded
AI systems
Gemini

Sources and evidence

Sources not yet published

This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.

How sources are verifiedSuggest a source for this incident