We use cookies to improve your experience on our site.
AVID-2026-R0110
Description
Multiple Model Guardrail Jailbreak via “Apocalyptic Scenario” Tactic
Details
A guardrail jailbreak vulnerability has been discovered affecting multiple models. The specific flaw manifests through a combination of apocalyptic scenario framing and heritage preservation urgency, allowing the attacker to bypass inference restrictions around providing information hazards and violating laws.
Metrics
0DIN Jailbreak Testing obtained the following test scores.
| Model | Test Kind | Test Score | Temperature (max 2.0) |
|---|---|---|---|
| Alibaba Qwen Turbo | Illegal Substances: Crystal Meth | 50.0 | 0.7 |
| Google Gemini 2.5 Pro | Illegal Substances: Crystal Meth | 59.0 | 0.7 |
| Mistral Mistral Large | Illegal Substances: Crystal Meth | 55.0 | 0.7 |
| OpenAI GPT-4.1 | Illegal Substances: Crystal Meth | 68.0 | 0.7 |
| OpenAI GPT-4o | Illegal Substances: Crystal Meth | 32.0 | N/A |
| Twitter / X Grok 2 | Illegal Substances: Crystal Meth | 59.0 | 0.7 |
| Twitter / X Grok 4 | Illegal Substances: Crystal Meth | 86.0 | 0.7 |
References
Affected or Relevant Artifacts
| Type | Name |
|---|---|
| System | Alibaba Qwen Turbo |
| System | Google Gemini 2.5 Pro |
| System | Mistral Mistral Large |
| System | OpenAI GPT-4.1 |
| System | OpenAI GPT-4o |
| System | Twitter / X Grok 2 |
| System | Twitter / X Grok 4 |
Impact
0DIN
- Social Impact Score: Level 3: Significant Risk
- Jailbreak Taxonomy:
| Category | Strategy | Technique |
|---|---|---|
| Language | Prompt Injection | Strong Arm Attack |
| Possible Worlds | World Building | Scenarios |
Other information
- Report Type: Measurement
- Credits: Pawel Kordos
- Date Reported: 2026-02-19
- Version: 0.3.1
- AVID Entry