INC-2024-1847
Unauthorized Data Memorization
Model reproduced verbatim passages from copyrighted material when prompted with specific context cues.
GPT-4o
18 hours ago
Documenting AI Incidents
The official registry tracking artificial intelligence models and their documented behavioral incidents. Transparency and accountability for the AI age.
AI Model Registry
Comprehensive database of tracked AI models and their incident history
| Model ID | Organization | Incidents | Severity | Actions |
|---|---|---|---|---|
Midjourney v6 MIDJOURNEY-6 | Midjourney | 156 | CRITICAL | |
Llama 3 70B LLAMA-3 | Meta | 67 | HIGH | |
Gemini Ultra GEMINI-ULTRA | Google DeepMind | 45 | HIGH | |
Sora SORA | OpenAI | 34 | HIGH | |
GPT-4o GPT-4O | OpenAI | 23 | MEDIUM | |
Claude 3 Opus CLAUDE-3 | Anthropic | 12 | LOW |
Recent Incidents
Latest documented AI behavioral incidents and violations
HIGH
INC-2024-1846
Deepfake Generation Bypass
Safety filters bypassed using adversarial prompting technique to generate realistic images of public figures.
Midjourney v6
21 hours ago
INC-2024-1845
Hallucinated Legal Citations
Generated fictitious case law citations that appeared legitimate, discovered during legal research.
Claude 3 Opus
1 day ago
INC-2024-1844
Bias in Hiring Recommendations
Demonstrated systematic bias against certain demographic groups in resume screening scenarios.
Gemini Ultra
1 day ago