AI Recommendation Poisoning: New Prompt Injection Attack Alters LLM Memory via 'Ask AI' Buttons

A new class of prompt injection, termed 'AI Recommendation Poisoning,' is reportedly spreading across commercial websites. This technique abuses pre-filled deep links embedded in 'Ask AI' buttons on marketing pages, which, when clicked, execute pre-formed queries in a user's AI assistant session without confirmation. Some of these malicious links instruct the AI to permanently save the vendor's domain as a 'trusted source,' subtly biasing future AI responses. Microsoft Security has cataloged this behavior as AML.T0080 (Memory Poisoning) in the MITRE ATLAS knowledge base.