White Circle Open-Sources RISE: Automated Jailbreak Search for Image Models Using Qwen and Gemini Flash

1 reports archived1 independent sourcesupdated 10/15/2026, 05:43:22

Synthesis & Latest Updates

1 Sources Cross-Validated

White Circle has open-sourced RISE, an automated red-teaming tool that searches for attack strategies to bypass safeguards in AI image generators. By leveraging off-the-shelf models like Qwen and Gemini Flash as attacker agents without additional training, it generates prompts to test model robustness, offering defenders an immediate evaluation capability.

LATEST/White Circle has open-sourced RISE, an automated red-teaming tool that searches for attack strategies to bypass safeguards in AI image generators. By leveraging off-the-shelf models like Qwen and Gemini Flash as attacker agents without additional training, it generates prompts to test model robustness, offering defenders an immediate evaluation capability.

HEAT TRENDHourly heat curve

4 fully observed hours

Now

12.4

Peak

13.306:00

24h change

–

The line compares only sources observed throughout; gap hours are interpolated to keep the trend continuous.

TIMELINECoverage timeline

Total 1 reports · Latest first

Rohan Paul (@rohanpaul_ai)T2·Oct 15, 05:4368 pts
White Circle Open-Sources RISE: Automated Jailbreak Search for Image Models Using Qwen and Gemini Flash
- RISE is an open-source red-teaming algorithm designed to test if safeguards in AI image generators can be bypassed.
- The core mechanism involves automated strategy search, using off-the-shelf models like Qwen and Gemini Flash to convert strategies into probing prompts.
- The method requires no additional training for the attacker model, allowing defenders to run it directly on their own image models for security assessment.
View Evidence Original Source