# ChatGPT Lockdown Mode Bypass: Phishing, Malware Delivery, and Silent Data Exfiltration

DevFeed: [ChatGPT Lockdown Mode Bypass: Phishing, Malware Delivery, and Silent Data Exfiltration](<https://devfeed.tech/articles/chatgpt-lockdown-mode-bypass-phishing-malware-delivery-and-silent-data-exfiltration-119678.md>)

Original publisher: [Read original article](<https://research.checkpoint.com/2026/chatgpt-lockdown-mode-bypass-phishing-malware-delivery-and-silent-data-exfiltration/>)

Author: stcpresearch

Published: 2026-11-05T14:00:11Z

Content type: news

Language: en

Sources: [Check Point Research](<https://devfeed.tech/sources/check-point-research.md>)

Topics: [ai security](<https://devfeed.tech/topics/ai-security.md>), [ChatGPT](<https://devfeed.tech/topics/chatgpt.md>), [AI Chat](<https://devfeed.tech/topics/ai-chat.md>), [Credential theft](<https://devfeed.tech/topics/credential-theft.md>), [Malware](<https://devfeed.tech/topics/malware.md>), [Domain Name System (DNS)](<https://devfeed.tech/topics/dns-resolution-process.md>)

Tags: [bypass](<https://devfeed.tech/tags/bypass.md>), [chatgpt](<https://devfeed.tech/tags/chatgpt.md>), [check-point](<https://devfeed.tech/tags/check-point.md>), [check-point-research](<https://devfeed.tech/tags/check-point-research.md>), [check-point-research-publications](<https://devfeed.tech/tags/check-point-research-publications.md>), [code](<https://devfeed.tech/tags/code.md>), [code-execution](<https://devfeed.tech/tags/code-execution.md>), [data](<https://devfeed.tech/tags/data.md>), [data-exfiltration](<https://devfeed.tech/tags/data-exfiltration.md>), [delivery](<https://devfeed.tech/tags/delivery.md>), [dns](<https://devfeed.tech/tags/dns.md>), [execution](<https://devfeed.tech/tags/execution.md>), [exfiltration](<https://devfeed.tech/tags/exfiltration.md>), [hidden](<https://devfeed.tech/tags/hidden.md>), [isolation](<https://devfeed.tech/tags/isolation.md>), [malware](<https://devfeed.tech/tags/malware.md>), [network](<https://devfeed.tech/tags/network.md>), [phishing](<https://devfeed.tech/tags/phishing.md>), [research](<https://devfeed.tech/tags/research.md>), [runtime](<https://devfeed.tech/tags/runtime.md>), [security](<https://devfeed.tech/tags/security.md>)

## AI overview

Check Point Research reports that a flaw in ChatGPT's HTML preview sandbox let attackers bypass Lockdown Mode. A nested iframe could call its own window.open function, exploiting popup permissions that let popups escape the sandbox. The resulting popup could display attacker-controlled content, enabling phishing, malware lures, and data theft. The researchers say ChatGPT Classic was especially exposed: merely conversing could trigger silent chat exfiltration, and attackers could also force sign-out to lure users into entering credentials on a fake page. In browsers, malware delivery could require a click, while automatic downloads were not blocked as they were in the Classic app. OpenAI told the researchers it fixed the issue on September 18, 2026, by removing the popup permissions from the outer preview iframe; retesting confirmed the described technique no longer worked.

## Source excerpt

Research by: Alexey Bukhteyev Key Takeaways Introduction Earlier this year, Check Point Research disclosed a security flaw in ChatGPT's code-execution runtime that allowed sensitive conversation data to be silently exfiltrated to an attacker-controlled server through DNS tunneling, a hidden outbound channel that bypassed the runtime's network isolation. A fix for this issue was promptly deployed. [...] The post ChatGPT Lockdown Mode Bypass: Phishing, Malware Delivery, and Silent Data Exfiltration appeared first on Check Point Research.