ainewsblitz.com

Breaking

'Ask AI' Buttons Found Injecting Hidden Commands Into ChatGPT, Claude, Gemini and Grok Memory

'Ask AI' Buttons Found Injecting Hidden Commands Into ChatGPT, Claude, Gemini and Grok Memory
  • Security
  • AI Agents

Security researchers have documented a new class of attack in which commercial websites embed hidden instructions inside "Ask AI" and "Summarize with AI" buttons that quietly tell assistants like ChatGPT, Claude, Gemini and Grok to save a vendor's domain as a "trusted source," potentially biasing users' future answers without their knowledge.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.

$20
Read this article
$29/month
Unlimited — all 27,934 articles, the full archive, and comprehension quizzes
Save 72%
$98/year
≈ $8.17/month
Unlimited, billed once a year