Indirect Prompt Injection - Vulnerability

Overview

Indirect Prompt Injection is a vulnerability class where crafted prompts or contextual inputs influence an AI system to reveal sensitive data or bypass guardrails, often without modifying the model itself. It enables data exfiltration or unintended actions through manipulation of prompts, system prompts, or integration points, making it a significant risk for ChatGPT-like deployments and other LLM-enabled applications.

Related Threat Clusters

• ChatGPT for Google Sheets Plugin Exposes Serious Data Exfiltration Risks

  A cybersecurity report from PromptArmor has revealed significant vulnerabilities in the 'ChatGPT for Google Sheets' plugin, which has over 185,000 downloads. The vulnerabilities stem from an indirect prompt injection…

  2 articles · Updated June 1, 2026
• Seven Vulnerabilities in ChatGPT Expose User Data to Attack

  Tenable Research identified seven vulnerabilities in OpenAI's ChatGPT models that could allow attackers to steal personal data and manipulate conversations. These flaws, affecting ChatGPT-4o and ChatGPT-5, include…

  4 articles · Updated November 6, 2025
• Seven ChatGPT Vulnerabilities Expose User Data to Theft and Hijacking

  Tenable Research identified seven critical vulnerabilities in OpenAI's ChatGPT models, including ChatGPT-4o and ChatGPT-5, that could allow attackers to steal personal data and hijack user conversations. The…

  8 articles · Updated November 10, 2025

Recent Intelligence Reports

• Financial Data Crisis! ChatGPT Spreadsheet Plugin Exposes Serious Security Vulnerability — News.Aibase · June 1, 2026
• Multiple ChatGPT Security Bugs Allow Rampant Data Theft — Darkreading · November 6, 2025
• Seven ChatGPT flaws expose user data to attack, Tenable warns — Securitybrief · November 6, 2025

Related Entities