OpenAI 已修補漏洞,Check Point 亦確認相關手法不再奏效;事件提醒大家,AI 安全除了模型防護,也要涵蓋處理不可信內容的應用程式功能。[1][2]
What did Check Point Research discover about the ChatGPT Lockdown Mode bypass that OpenAI patched on September 18, 2026, how did Lockdown MoThe reported bypass used ChatGPT’s HTML preview sandbox, an application feature separate from the model’s safety controls.
AI 提示
Create a landscape editorial hero image for this Studio Global article: What did Check Point Research discover about the ChatGPT Lockdown Mode bypass that OpenAI patched on September 18, 2026, how did Lockdown Mo. Article summary: Check Point Research found an application-level weakness in ChatGPT’s HTML-preview sandbox that bypassed Lockdown Mode, enabling silent chat-data theft, phishing and malware delivery rather than defeating the model’s saf. Topic tags: general, general web. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text, watermarks, charts with fake numbers, clic
openai.com
Check Point Research 發現,ChatGPT 顯示 HTML 預覽的沙盒曾可被利用,繞過 Lockdown Mode 的防護。研究人員指,漏洞可能令對話資料在用戶不知情下外傳,亦可用於釣魚和派送惡意軟件。12
OpenAI 在負責任披露後修補問題;Check Point 表示,重新測試後確認相關手法已經失效。12
OpenAI 已修補漏洞,Check Point 亦確認相關手法不再奏效;事件提醒大家,AI 安全除了模型防護,也要涵蓋處理不可信內容的應用程式功能。[1][2]
What did Check Point Research discover about the ChatGPT Lockdown Mode bypass that OpenAI patched on September 18, 2026, how did Lockdown MoThe reported bypass used ChatGPT’s HTML preview sandbox, an application feature separate from the model’s safety controls.
AI 提示
Create a landscape editorial hero image for this Studio Global article: What did Check Point Research discover about the ChatGPT Lockdown Mode bypass that OpenAI patched on September 18, 2026, how did Lockdown Mo. Article summary: Check Point Research found an application-level weakness in ChatGPT’s HTML-preview sandbox that bypassed Lockdown Mode, enabling silent chat-data theft, phishing and malware delivery rather than defeating the model’s saf. Topic tags: general, general web. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text, watermarks, charts with fake numbers, clic
openai.com
Check Point Research 發現,ChatGPT 顯示 HTML 預覽的沙盒曾可被利用,繞過 Lockdown Mode 的防護。研究人員指,漏洞可能令對話資料在用戶不知情下外傳,亦可用於釣魚和派送惡意軟件。12
OpenAI 在負責任披露後修補問題;Check Point 表示,重新測試後確認相關手法已經失效。12