Find notable cyber news and cases, enriched with sources, timelines, and signals.

OpenAI ShadowLeak mitigation for ChatGPT

Advisory/Mitigation
First reported
Last updated
Happening score
H score 18
2 unique sources, 2 articles

Summary

Hide ▲

OpenAI acknowledged and fixed ShadowLeak, ending a stealthy ChatGPT email-exfiltration path in the studied Gmail integration. The mitigation mattered because the flaw let hidden HTML instructions in email bodies steer the agent without visible network traces on the victim side. The key change was issuer-side closure of the attack path rather than a user-side response.

Related Happenings

Cline Kanban server WebSocket origin/authentication security flaw

Vulnerability
H score33 First: 07.05.2026 17:30 Last: 07.05.2026 17:30 Sources 1

About this happening: Cline Kanban server has a critical WebSocket origin/authentication flaw that can let a webpage a developer visits exfiltrate workspace data, inject terminal commands...

LeRobot unsafe deserialization RCE (CVE-2026-25874)

Vulnerability
H score33 First: 28.04.2026 14:18 Last: 28.04.2026 14:18 Sources 1

About this happening: The CVE-2026-25874 unsafe deserialization flaw in Hugging Face LeRobot enables unauthenticated remote code execution against network-reachable deployments. The bug sit...

Microsoft SharePoint Server spoofing vulnerability (actively exploited) (CVE-2026-32201)

Vulnerability
H score55 First: 14.04.2026 20:41 Last: 14.04.2026 20:41 Sources 1

About this happening: Microsoft patched CVE-2026-32201 in Microsoft SharePoint Server, a spoofing vulnerability that was exploited in attacks and could affect confidentiality and ...

ChatGPT single-prompt DNS side-channel exfiltration remote code execution flaw

Vulnerability
H score33 First: 31.03.2026 16:01 Last: 31.03.2026 16:01 Sources 1

About this happening: A ChatGPT vulnerability let a single malicious prompt covertly exfiltrate prompts, messages, uploaded files, and other sensitive content through a DNS side channel. Th...

OpenAI Safety Bug Bounty launch

Commercial Activity
H score0 First: 26.03.2026 14:20 Last: 26.03.2026 14:20 Sources 1

About this happening: OpenAI launched the Safety Bug Bounty on Bugcrowd, expanding researcher coverage for AI abuse and safety risks across its products. The new program complements...

Timeline

  1. 19.09.2025 22:07 2 articles · 9mo ago

    OpenAI ShadowLeak mitigation for ChatGPT

    Initial Disclosure

    After the issue was reported, the studied ChatGPT/Gmail attack path stopped working by August, indicating OpenAI had already mitigated ShadowLeak. That early closure mattered because it removed the stealthy email-exfiltration path in the test environment.

    Show sources