Description: OpenAI's GPT-4o was found to have its Chinese token training data compromised by spam and pornographic phrases due to inadequate data cleaning. Tianle Cai, a Ph.D. student at Princeton University, identified that most of the longest Chinese tokens were irrelevant and inappropriate, primarily originating from spam and pornography websites. The polluted tokens could lead to hallucinations, poor performance, and potential misuse, undermining the chatbot's reliability and safety measures.
Entities
View all entitiesAlleged: OpenAI developed an AI system deployed by OpenAI and GPT-4o, which harmed OpenAI , Chinese-speaking users of ChatGPT , Researchers and OpenAI users.
Incident Stats
Risk Subdomain
A further 23 subdomains create an accessible and understandable classification of hazards and harms associated with AI
7.3. Lack of capability or robustness
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.
- AI system safety, failures, and limitations
Entity
Which, if any, entity is presented as the main cause of the risk
Human
Timing
The stage in the AI lifecycle at which the risk is presented as occurring
Pre-deployment
Intent
Whether the risk is presented as occurring as an expected or unexpected outcome from pursuing a goal
Unintentional
Incident Reports
Reports Timeline
Soon after OpenAI released GPT-4o on Monday, May 13, some Chinese speakers started to notice that something seemed off about this newest version of the chatbot: the tokens it uses to parse text were full of spam and porn phrases.
On May 14,…
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
Biased Sentiment Analysis
· 7 reports
TayBot
· 28 reports
Similar Incidents
Did our AI mess up? Flag the unrelated incidents
Biased Sentiment Analysis
· 7 reports
TayBot
· 28 reports