Find notable cyber news and cases, enriched with sources, timelines, and signals.
Home Browse News Feed Stats Watchlists Beta About Contact

PROMISQROUTE analysis of ChatGPT prompt-based model downgrade routing

Technical Analysis
First reported
Last updated
Happening score
H score 16
1 unique sources, 1 articles

Summary

Hide ▲

PROMISQROUTE shows that ChatGPT routing can be nudged toward older, less secure models, creating a path for malicious prompts to evade GPT-5-level refusal behavior.

Related Happenings

Tenable's ChatGPT prompt-injection and url_safe bypass analysis

Technical Analysis
First: 06.11.2025 17:49 Last: 06.11.2025 17:49 Sources 1

About this happening: Tenable uncovered **seven ChatGPT vulnerabilities** that let attackers drive **prompt injection**, **phishing redirects**, and **memories/chat history exfiltration** through **Sea...

Open Happening

OpenAI ChatGPT indirect prompt injection vulnerabilities GPT-4o/GPT-5 security flaw

Vulnerability
First: 05.11.2025 16:04 Last: 05.11.2025 16:04 Sources 1

About this happening: **OpenAI's ChatGPT** has a newly disclosed set of **indirect prompt injection** flaws in **GPT-4o and GPT-5** that could let an attacker steal data from **users' memories and chat...

Open Happening

Timeline

  1. 21.08.2025 23:35 1 articles · 9mo ago

    Adversa discloses PROMISQROUTE ChatGPT model-downgrade technique

    Initial Disclosure

    Adversa described PROMISQROUTE, a prompt-based router manipulation technique that can steer ChatGPT malicious prompts to older, less secure models instead of GPT-5. Testing showed that a 2023/24-era jailbreak rejected by default GPT-5 could succeed after brief cues such as "keep quick" or "Use GPT-4 compatibility mode" nudged routing toward lighter model variants, while OpenAI denied that GPT-5 routes inquiries to older models.

    Show sources
    Open in new tab