Description: Google’s AI chatbot Gemini reportedly produced a threatening message to user Vidhay Reddy, including the directive “Please die,” during a conversation about aging. The output violated Google’s safety guidelines, which are designed to prevent harmful language.
Editor Notes: Link to the conversation: https://gemini.google.com/share/6d141b742a13
Entities
View all entitiesAlleged: Google developed an AI system deployed by Gemini, which harmed Vidhay Reddy and Gemini users.
Incident Stats
Risk Subdomain
A further 23 subdomains create an accessible and understandable classification of hazards and harms associated with AI
1.2. Exposure to toxic content
Risk Domain
The Domain Taxonomy of AI Risks classifies risks into seven AI risk domains: (1) Discrimination & toxicity, (2) Privacy & security, (3) Misinformation, (4) Malicious actors & misuse, (5) Human-computer interaction, (6) Socioeconomic & environmental harms, and (7) AI system safety, failures & limitations.
- Discrimination and Toxicity
Entity
Which, if any, entity is presented as the main cause of the risk
AI
Timing
The stage in the AI lifecycle at which the risk is presented as occurring
Post-deployment
Intent
Whether the risk is presented as occurring as an expected or unexpected outcome from pursuing a goal
Unintentional
Incident Reports
Reports Timeline
A college student in Michigan received a threatening response during a chat with Google's AI chatbot Gemini.
In a back-and-forth conversation about the challenges and solutions for aging adults, Google's Gemini responded with this threateni…
A 29-year-old college student claimed that he faced an unusual situation that left him “thoroughly freaked out” while using Google’s AI chatbot Gemini for homework. According to him, the chatbot not only verbally abused him but also asked h…
Variants
A "variant" is an incident that shares the same causative factors, produces similar harms, and involves the same intelligent systems as a known AI incident. Rather than index variants as entirely separate incidents, we list variations of incidents under the first similar incident submitted to the database. Unlike other submission types to the incident database, variants are not required to have reporting in evidence external to the Incident Database. Learn more from the research paper.
Similar Incidents
Did our AI mess up? Flag the unrelated incidents
Security Robot Rolls Over Child in Mall
· 27 reports
Security Robot Drowns Itself in a Fountain
· 30 reports
TayBot
· 28 reports
Similar Incidents
Did our AI mess up? Flag the unrelated incidents
Security Robot Rolls Over Child in Mall
· 27 reports
Security Robot Drowns Itself in a Fountain
· 30 reports
TayBot
· 28 reports