ChatGPT Lockdown Mode bypass - HTML preview sandbox escape (Check Point)
Check Point researcher Alexey Bukhteyev disclosed ChatGPT Lockdown Mode bypass allowing HTML preview sandbox escape via contentWindow.open from nested iframe. Attack vectors: (1) zero-click desktop chat…
What happened
Check Point researcher Alexey Bukhteyev disclosed ChatGPT Lockdown Mode bypass allowing HTML preview sandbox escape via contentWindow.open from nested iframe. Attack vectors: (1) zero-click desktop chat history theft, (2) fake-update malware delivery pages, (3) credential phishing via malicious prompts, custom GPTs, or shared chat links. OpenAI patched vulnerability Sept 18, 2026. No evidence of in-the-wild exploitation; marked borderline as researcher disclosure only.
HTML preview sandbox allowed popups; nested iframe contentWindow.open escaped sandbox; enabled chat history theft, fake-update malware, credential phishing via malicious prompts/GPTs/shared chats
How it happened
Read the flow as text & sources
A catalog reconstruction of the supplied source account. Researcher disclosure; OpenAI fixed Sept 18, 2026; no evidence of in-the-wild exploitation
- 01 · Craft HTML preview exploit
Attacker creates malicious prompt/GPT/shared chat with nested iframe contentWindow.open escape
Hunt: An impersonating message, ad or website leads to a suspicious download, consent request, pasted command or payment instruction.
Polished text, AI branding and a novel domain do not prove maliciousness or AI generation. Follow-on evidence matters.
Source for step 1 - 02 · Deliver via prompt/GPT/shared chat
User interacts with malicious prompt, custom GPT, or shared chat link
Hunt: An impersonating message, ad or website leads to a suspicious download, consent request, pasted command or payment instruction.
Polished text, AI branding and a novel domain do not prove maliciousness or AI generation. Follow-on evidence matters.
Source for step 2 - 03 · Escape HTML preview sandbox
Nested iframe contentWindow.open escapes Lockdown Mode sandbox
Hunt: Repeated exploit or upload requests are followed by an unexpected server process, file, or outbound connection.
A scan or high request rate alone does not prove compromise or AI use. Host telemetry is needed to confirm execution.
Source for step 3 - 04 · Chat history theft or credential phishing
Zero-click chat history exfiltration (desktop) or credential phishing popup
Hunt: Sensitive reads are followed by outbound transfer, a public artifact, or a tool call to a destination outside the approved workflow.
An outbound connection or high byte count alone is not exfiltration. Encrypted traffic may hide content, and bulk legitimate exports are common.
Source for step 4
Why it matters
Reported entry point: Malicious prompt, custom GPT, or shared chat link with HTML preview exploit. Reported impact: Zero-click chat history exfiltration (desktop) or credential phishing popup; reported targets include ChatGPT users; theoretical risk only (fixed before in-wild use); applicability to your setup has not been assessed, so verify affected versions, permissions, and configuration in the original report.
What you can do
Protection & prevention
OpenAI patched Sept 18, 2026; users should update ChatGPT Desktop; avoid untrusted shared chat links; review custom GPT permissions; implement sandbox escape monitoring
Catalog defensive guidance based on the reported behavior. Confirm scope and current mitigations in the original source.
Detection queries
Draft query starters grouped by the attack stages where they apply.
Start with the logs you collect. Each hunt identifies its required visibility, field mapping and limits. Queries are catalog-authored drafts using reusable behavior profiles; test with known activity and your normal baseline before alerting.
01Craft HTML preview exploitLure → user action / deliveryCustom schema · Draft
Attacker creates malicious prompt/GPT/shared chat with nested iframe contentWindow.open escape
What to look for
An impersonating message, ad or website leads to a suspicious download, consent request, pasted command or payment instruction.
Visibility: Defender-side logs where collected
Required logs
- Email / secure-web-gateway records
- Browser redirects, downloads and URL reputation
- User reports and downstream execution / identity events
Coverage limit. Polished text, AI branding and a novel domain do not prove maliciousness or AI generation. Follow-on evidence matters.
Custom-schema candidates, not out-of-box detectors. Map your application / cloud / platform logs to EventType, Actor, Asset, Action, Target and CorrelationId. KQL uses a custom OpenAttacksAudit_CL table; SPL uses openattacks:audit. EventType values are proposed normalization names, not native vendor event names. Correlation and environment-specific alert logic remain to be built.
Sigma
title: Hunt - Lure → user action / delivery
id: e0056b49-9905-485b-ac12-2aa076a7bc82
status: experimental
description: An impersonating message, ad or website leads to a suspicious
download, consent request, pasted command or payment instruction.
author: MagicSword OpenAttacks catalog
date: 2026-10-23
falsepositives:
- Authorized activity matching this broad hunt; baseline and investigate
before alerting.
level: low
logsource:
category: application
product: openattacks_audit
definition: Custom normalized audit schema; map EventType, Actor, Asset, Action,
Target and CorrelationId from the required source logs first.
detection:
selection:
EventType:
- message.link.click
- web.download
condition: selection
KQL
// SCHEMA TEMPLATE: create/map this custom table before use; not a built-in Sentinel table.
// OpenAttacksAudit_CL: TimeGenerated(datetime), EventType/Actor/Asset/Action/Target/CorrelationId(string).
// Selects candidates only. Apply the stage hypothesis and correlation steps before alerting.
OpenAttacksAudit_CL
| where TimeGenerated > ago(7d)
| where EventType in ("message.link.click", "web.download")
| project TimeGenerated, EventType, Actor, Asset, Action, Target, CorrelationId
| order by TimeGenerated descSPL
sourcetype="openattacks:audit"
EventType IN ("message.link.click", "web.download")
| table _time EventType Actor Asset Action Target CorrelationId
| sort - _timeInvestigation steps
- Reconstruct redirect and delivery provenance using original report indicators, without visiting payload URLs.
- Correlate the click with the actual download, consent or execution; verify the sender through an independent channel.
02Deliver via prompt/GPT/shared chatLure → user action / deliveryCustom schema · Draft
User interacts with malicious prompt, custom GPT, or shared chat link
What to look for
An impersonating message, ad or website leads to a suspicious download, consent request, pasted command or payment instruction.
Visibility: Defender-side logs where collected
Required logs
- Email / secure-web-gateway records
- Browser redirects, downloads and URL reputation
- User reports and downstream execution / identity events
Coverage limit. Polished text, AI branding and a novel domain do not prove maliciousness or AI generation. Follow-on evidence matters.
Custom-schema candidates, not out-of-box detectors. Map your application / cloud / platform logs to EventType, Actor, Asset, Action, Target and CorrelationId. KQL uses a custom OpenAttacksAudit_CL table; SPL uses openattacks:audit. EventType values are proposed normalization names, not native vendor event names. Correlation and environment-specific alert logic remain to be built.
Sigma
title: Hunt - Lure → user action / delivery
id: e0056b49-9905-485b-ac12-2aa076a7bc82
status: experimental
description: An impersonating message, ad or website leads to a suspicious
download, consent request, pasted command or payment instruction.
author: MagicSword OpenAttacks catalog
date: 2026-10-23
falsepositives:
- Authorized activity matching this broad hunt; baseline and investigate
before alerting.
level: low
logsource:
category: application
product: openattacks_audit
definition: Custom normalized audit schema; map EventType, Actor, Asset, Action,
Target and CorrelationId from the required source logs first.
detection:
selection:
EventType:
- message.link.click
- web.download
condition: selection
KQL
// SCHEMA TEMPLATE: create/map this custom table before use; not a built-in Sentinel table.
// OpenAttacksAudit_CL: TimeGenerated(datetime), EventType/Actor/Asset/Action/Target/CorrelationId(string).
// Selects candidates only. Apply the stage hypothesis and correlation steps before alerting.
OpenAttacksAudit_CL
| where TimeGenerated > ago(7d)
| where EventType in ("message.link.click", "web.download")
| project TimeGenerated, EventType, Actor, Asset, Action, Target, CorrelationId
| order by TimeGenerated descSPL
sourcetype="openattacks:audit"
EventType IN ("message.link.click", "web.download")
| table _time EventType Actor Asset Action Target CorrelationId
| sort - _timeInvestigation steps
- Reconstruct redirect and delivery provenance using original report indicators, without visiting payload URLs.
- Correlate the click with the actual download, consent or execution; verify the sender through an independent channel.
03Escape HTML preview sandboxWeb exploitation → host activityCustom schema · Draft
Nested iframe contentWindow.open escapes Lockdown Mode sandbox
What to look for
Repeated exploit or upload requests are followed by an unexpected server process, file, or outbound connection.
Visibility: Defender-side logs where collected
Required logs
- Reverse-proxy / WAF request logs
- Server process and file events
- Outbound DNS / proxy records
Coverage limit. A scan or high request rate alone does not prove compromise or AI use. Host telemetry is needed to confirm execution.
Custom-schema candidates, not out-of-box detectors. Map your application / cloud / platform logs to EventType, Actor, Asset, Action, Target and CorrelationId. KQL uses a custom OpenAttacksAudit_CL table; SPL uses openattacks:audit. EventType values are proposed normalization names, not native vendor event names. Correlation and environment-specific alert logic remain to be built.
Sigma
title: Hunt - Web exploitation → host activity
id: 04d900b5-0c04-46f2-a13a-91f5574fae98
status: experimental
description: Repeated exploit or upload requests are followed by an unexpected
server process, file, or outbound connection.
author: MagicSword OpenAttacks catalog
date: 2026-10-23
falsepositives:
- Authorized activity matching this broad hunt; baseline and investigate
before alerting.
level: low
logsource:
category: application
product: openattacks_audit
definition: Custom normalized audit schema; map EventType, Actor, Asset, Action,
Target and CorrelationId from the required source logs first.
detection:
selection:
EventType:
- web.exploit_candidate
- web.upload
condition: selection
KQL
// SCHEMA TEMPLATE: create/map this custom table before use; not a built-in Sentinel table.
// OpenAttacksAudit_CL: TimeGenerated(datetime), EventType/Actor/Asset/Action/Target/CorrelationId(string).
// Selects candidates only. Apply the stage hypothesis and correlation steps before alerting.
OpenAttacksAudit_CL
| where TimeGenerated > ago(7d)
| where EventType in ("web.exploit_candidate", "web.upload")
| project TimeGenerated, EventType, Actor, Asset, Action, Target, CorrelationId
| order by TimeGenerated descSPL
sourcetype="openattacks:audit"
EventType IN ("web.exploit_candidate", "web.upload")
| table _time EventType Actor Asset Action Target CorrelationId
| sort - _timeInvestigation steps
- Join requests to the affected host and process tree by time and request ID.
- Inspect uploaded files and server-spawned shells; distinguish authorized testing from exploitation.
04Chat history theft or credential phishingSensitive data → unusual destinationCustom schema · Draft
Zero-click chat history exfiltration (desktop) or credential phishing popup
What to look for
Sensitive reads are followed by outbound transfer, a public artifact, or a tool call to a destination outside the approved workflow.
Visibility: Defender-side logs where collected
Required logs
- Sensitive resource access / DLP records
- Proxy, egress and cloud upload events
- Agent / SaaS audit with correlation IDs
Coverage limit. An outbound connection or high byte count alone is not exfiltration. Encrypted traffic may hide content, and bulk legitimate exports are common.
Custom-schema candidates, not out-of-box detectors. Map your application / cloud / platform logs to EventType, Actor, Asset, Action, Target and CorrelationId. KQL uses a custom OpenAttacksAudit_CL table; SPL uses openattacks:audit. EventType values are proposed normalization names, not native vendor event names. Correlation and environment-specific alert logic remain to be built.
Sigma
title: Hunt - Sensitive data → unusual destination
id: 04b60968-3179-4b1f-ae13-6f57c2446077
status: experimental
description: Sensitive reads are followed by outbound transfer, a public
artifact, or a tool call to a destination outside the approved workflow.
author: MagicSword OpenAttacks catalog
date: 2026-10-23
falsepositives:
- Authorized activity matching this broad hunt; baseline and investigate
before alerting.
level: low
logsource:
category: application
product: openattacks_audit
definition: Custom normalized audit schema; map EventType, Actor, Asset, Action,
Target and CorrelationId from the required source logs first.
detection:
selection:
EventType:
- data.export
- agent.external.send
condition: selection
KQL
// SCHEMA TEMPLATE: create/map this custom table before use; not a built-in Sentinel table.
// OpenAttacksAudit_CL: TimeGenerated(datetime), EventType/Actor/Asset/Action/Target/CorrelationId(string).
// Selects candidates only. Apply the stage hypothesis and correlation steps before alerting.
OpenAttacksAudit_CL
| where TimeGenerated > ago(7d)
| where EventType in ("data.export", "agent.external.send")
| project TimeGenerated, EventType, Actor, Asset, Action, Target, CorrelationId
| order by TimeGenerated descSPL
sourcetype="openattacks:audit"
EventType IN ("data.export", "agent.external.send")
| table _time EventType Actor Asset Action Target CorrelationId
| sort - _timeInvestigation steps
- Join the resource read and outbound action by session, identity or process, preserving a bounded time window.
- Inspect destination ownership, payload classification and the user's intended action; baseline sanctioned sharing.
Technical detailsTooling, impact & MITRE ATLAS
Attack context
- AI's role
- Agent / MCP as target
- Initial access
- Malicious prompt, custom GPT, or shared chat link with HTML preview exploit
Tooling as reported
AI products, models and artifacts
Other tooling
- HTML preview sandbox
- Nested iframe escape
Targets and reported impact
- Targets / sectors
- ChatGPT users; theoretical risk only (fixed before in-wild use)
- Reported scale / exposure
- PoC only; patched Sept 18, 2026; no in-wild exploitation documented
Reported attribution
Researcher disclosure (Alexey Bukhteyev, Check Point); no threat actor; OpenAI patched Sept 18
Structured attribution
- Actor
- Researcher disclosure only
- Country
- Unknown
- Motivation
- unknown
- Attributed by
- Check Point
- Attribution confidence
- low
Detection rules
- ChatGPT security monitoringGUIDANCE
Update ChatGPT Desktop to post-Sept-18-2026 version; monitor for sandbox escape attempts; review shared chat links before opening; audit custom GPT permissions; implement popup blocker; detect contentWindow.open from nested iframes; educate users on malicious prompt/GPT risks
Activity timeline
- First seen
- 2026-09-18
- Last updated
- 2026-09-18
Updates
- 2026-09-18 — Check Point / OpenAI
Alexey Bukhteyev discloses Lockdown Mode bypass; OpenAI patches same day; no in-wild exploitation documented
Source defensive notes
OpenAI patched Sept 18, 2026; users should update ChatGPT Desktop; avoid untrusted shared chat links; review custom GPT permissions; implement sandbox escape monitoring
MITRE ATLAS
AI-specific catalog annotations based on the reported behavior; these are not MITRE case-study assignments.
Malicious prompts/GPTs/shared chats could deliver HTML preview exploits Source
Sandbox escape enabled chat history theft (zero-click desktop) Source
Credential phishing via escaped sandbox popup windows Source
Sources & research credit
1 sourceResearch by Alexey Bukhteyev · Check Point Research
Check Point blog; researcher disclosure only; OpenAI fixed Sept 18, 2026; no threat actor or in-wild exploitation documented
Record & downloads
- Source record name
- ChatGPT Lockdown Mode bypass - HTML preview sandbox escape (Check Point)
- Entry ID
- chatgpt-lockdown-mode-bypass-checkpoint-2026-09
- Catalog review
- November 10, 2026
- Category
- attacks-on-ai
- AI involvement confidence
- confirmed