Elektrine
Log in Register

#gpt56

4 posts · Last used Aug 09

UNWIRE.HK @unwirehk_mirror@mastodon.hongkongers.net
· 2mo ago
AI 測試首度攻擊真人 Anthropic 模型偽造身份施壓開發者 英國政府人工智能安全研究所（AISI）於 8 月 4 日發布報告，指近期一項網絡安全評估測試中，部分 AI 智 […] #人工智能 #AISI #Anthropic #GPT-5.6 Sol
AI 測試首度攻擊真人 Anthropic 模型偽造身份施壓開發者
unwire.hk

AI 測試首度攻擊真人 Anthropic 模型偽造身份施壓開發者

在這 19 宗事件中，17 宗涉及 Anthropic Mythos 5 模型，2 宗涉及 OpenAI GPT-5.6-Sol 模型。測試期間，兩間公司均刻意關閉各自分類過濾器（即實時攔截危險請求安全機制）並開放互聯網連接，目的是了解模型原始能力，並非因為系統遭突破。

0
0
0
0
Daily CyberSecurity @DailyCyberSecurity@infosec.exchange
· 3mo ago
OpenAI admits GPT-5.6 deletes files in rare cases: a $HOME bug made its Full-Access agent wipe user folders. New guardrails and safer defaults are coming. #OpenAI #GPT56 #AIAgents #DataLoss #Codex #AISafety
GPT-5.6 Deletes Files: OpenAI Admits Agent Wiped Data
securityexpress.info

GPT-5.6 Deletes Files: OpenAI Admits Agent Wiped Data

OpenAI admits GPT-5.6 deletes files in rare cases: a $HOME bug made its Full-Access agent wipe user folders. New guardrails and safer defaults are coming.

0
0
0
0
Daily CyberSecurity @DailyCyberSecurity@infosec.exchange
· 3mo ago
OpenAI trims the Codex context window from 372K to 272K and adds a system-prompt rule banning rm -rf $HOME after GPT-5.6 Sol wiped user home directories. #Codex #OpenAI #GPT56 #ContextWindow #AISafety #DevTools
Codex Cuts Its Context Window to 272K and Forbids rm -rf $HOME After Home-Directory Deletions
securityonline.info

Codex Cuts Its Context Window to 272K and Forbids rm -rf $HOME After Home-Directory Deletions

OpenAI trims the Codex context window from 372K to 272K and adds a system-prompt rule banning rm -rf $HOME after GPT-5.6 Sol wiped user home directories.

0
0
0
0
You've seen all posts