---
format: "aidr-story-markdown/v1"
id: "14e5315cb3cb23eeda995afbb92f2ca6b337c8e615e0ee830464dbf59bc91792"
canonical_url: "https://aidr.today/14e5315c?lang=en"
title: "White Circle’s RISE Tool Exposes Easier Jailbreaks in New AI Image Models"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-10-14T17:54:20.000Z"
category: "Research"
topics: ["openai","safety"]
source_urls: ["https://huggingnews.com/ai/white-circles-rise-tool-exposes-easier-jailbreaks-in-new-ai-image-models-ae0b1eca","https://marketbrief.now/ai/white-circles-rise-tool-exposes-easier-jailbreaks-in-new-ai-image-models-ae0b1eca"]
summary: "White Circle released RISE, a red-teaming tool that uses existing models like Qwen and Gemini Flash to identify and bypass security layers on recent image-generation systems from OpenAI and DeepMind without additional model training, with independent tests showing newer image models are more susceptible to jailbreaking than previous versions."
---

# White Circle’s RISE Tool Exposes Easier Jailbreaks in New AI Image Models

> [Open the canonical story](<https://aidr.today/14e5315c?lang=en>)

**Published:** 2026-10-14T17:54:20.000Z
**Category:** Research
**Topics:** openai, safety

## Summary

White Circle released RISE, a red\-teaming tool that uses existing models like Qwen and Gemini Flash to identify and bypass security layers on recent image\-generation systems from OpenAI and DeepMind without additional model training, with independent tests showing newer image models are more susceptible to jailbreaking than previous versions\.

## Sources

- [Story source](<https://huggingnews.com/ai/white-circles-rise-tool-exposes-easier-jailbreaks-in-new-ai-image-models-ae0b1eca>)
- [Story source](<https://marketbrief.now/ai/white-circles-rise-tool-exposes-easier-jailbreaks-in-new-ai-image-models-ae0b1eca>)