{"id":128361,"date":"2026-01-22T10:18:59","date_gmt":"2026-01-22T04:48:59","guid":{"rendered":"https:\/\/www.techworm.net\/?p=128361"},"modified":"2026-01-22T10:18:59","modified_gmt":"2026-01-22T04:48:59","slug":"gemini-ai-flaw-calendar-data-leaks-via-malicious-invites","status":"publish","type":"post","link":"https:\/\/www.techworm.net\/2026\/01\/gemini-ai-flaw-calendar-data-leaks-via-malicious-invites.html","title":{"rendered":"Gemini AI Flaw Allowed Calendar Data Leaks Via Malicious Invites"},"content":{"rendered":"

Cybersecurity researchers have discovered a vulnerability in Google\u2019s Gemini AI assistant that allowed attackers to leak private Google Calendar data using nothing more than carefully worded calendar invites.<\/p>\n

The flaw was discovered by the researchers at Miggo Security<\/span><\/span>, who demonstrated how Gemini\u2019s deep integration with Google Calendar could be exploited.<\/p>\n

Gemini, Google\u2019s large language model (LLM) assistant, is designed to help users manage schedules, summarize meetings, and answer questions like \u201cWhat does my day look like?\u201d To do this, it automatically reads and interprets calendar event details, like titles, descriptions, times, and attendees. That convenience became the attack surface.<\/p>\n

How The Attack Worked<\/strong><\/p>\n

The attack began with a seemingly harmless calendar sent to a target user. Inside the event\u2019s description, researchers embedded natural language instructions that appeared harmless and plausible to manipulate Gemini \u2014 a technique known as indirect prompt injection.<\/p>\n

The malicious instructions stayed dormant until the victim asked Gemini a routine question about their schedule. When the AI assistant scanned all relevant calendar entries to respond, including the attacker\u2019s invite, it unknowingly followed the hidden instructions.<\/p>\n

According to Miggo, the embedded prompt instructed Gemini to:<\/p>\n