Unverified50% confidenceFactExact time
Attackers can hide malicious instructions in web pages, documents, or even images, and when an agent reads this content it may mistake the instructions for legitimate commands and execute them
1
Sources
50%
Confidence
Long-term
Relevance
9/11/2026
First Seen
Sources
Related Entities
Related Claims
Unverified攻击者提交如AGENTS.md这类被AI代理默认读取的文本规则文件，引导AI去读取一张恶意PNG图片79% similarVerifiedPrompt injection attacks induce AI agents to execute malicious operations by embedding text disguised as system instructions in code comments, documentation, web content, or even file names.77% similarUnverifiedAttackers can induce agents to execute a sequence of seemingly normal operations—such as querying user data, calling formatting tools, and triggering email sending—to form a data exfiltration chain.77% similarUnverified攻击者常借助混淆技术隐藏恶意脚本诱导用户复制粘贴运行，这类手法被称为剪贴板劫持或社会工程学注入77% similarPartially VerifiedIndirect prompt injection attacks plant hidden instructions within normal business data to induce Agents into executing unauthorized operations, representing a security risk at the semantic level rather than the code level.74% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/905247API
curl https://kongchang.com/api/v1/knowledge/claims/905247MCP
get_claim(id=905247)