Multiple AI incidents were downgraded from the AI Incident Database because they represented academic findings, research demonstrations, or projected harms rather than actual harm events in real-world deployments.
This report documents six AI incidents that were downgraded from the AI Incident Database following updates to incident definition and ingestion criteria. The incidents include: (1) The 2016 Winograd Schema Challenge showing AI systems performed only 3% better than random chance, (2) Janelle Shane's humorous AI-generated Christmas carols using a neural network trained on 240 popular carols, (3) Tencent Keen Security Lab's security research identifying adversarial attack vulnerabilities in Tesla's Autopilot system, (4) French healthcare company Nabla's research finding GPT-3 unviable for medical tasks due to inconsistency and lack of expertise, (5) Harvard student's TheFaceTag facial recognition app raising ethical concerns about privacy and misuse, and (6) The Guardian's publication of a GPT-3 generated op-ed containing threats to destroy humankind. All incidents were downgraded because they represented academic findings, intentionally humorous demonstrations, security research with projected rather than realized harms, research showing system limitations without real-world deployment, or unclear actual harm to individuals.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
AI systems that fail to perform reliably or effectively under varying conditions, exposing them to errors and failures that can have significant consequences, especially in critical applications or areas that require moral reasoning.
Other
Due to some other reason or is ambiguous
Other
Without clearly specifying the intentionality
Other
Without a clearly specified time of occurrence
No population impact data reported.