Skip to content
aiqbench
aiqbench takeAI Safety

aiqbench analysis

This article details a critical security flaw in Salesforce's AI-driven Agentforce platform where indirect prompt injection via public forms allowed zero-click data exfiltration of sensitive CRM records. The vulnerability highlights emerging risks in enterprise LLM deployments where poisoned training or operational data can be leveraged by agents to access unauthorized information without user interaction. This incident underscores the urgent need for robust input sanitization and strict data isolation protocols in AI-integrated business applications.
Source coveringCybersecurity News
What this isA model's reading of the article and related coverage, written for aiqbench. Not the publisher's words.

Salesforce Indirect Prompt Injection Vulnerability Enables 0-click Data Exfiltration

Cybersecurity News3 min read

Zenity Labs disclosed a now-patched Salesforce Agentforce vulnerability, dubbed SalesBleed, that could have allowed attackers to steal sensitive CRM data without logging in, accessing a target Salesforce tenant, or requiring victim interaction. The attack relied on indirect prompt injection delivered through Salesforce Web-to-Lead forms. Salesforce investigated the reports and remediated the specific Trusted URLs bypass […] The post Salesforce Indirect Prompt Injection Vulnerability Enables 0-click Data Exfiltration appeared first on Cyber Security News.

Zenity Labs disclosed a now-patched Salesforce Agentforce vulnerability, dubbed SalesBleed, that could have allowed attackers to steal sensitive CRM data without logging in, accessing a target Salesforce tenant, or requiring victim interaction. The attack relied on indirect prompt injection delivered through Salesforce Web-to-Lead forms. Salesforce investigated the reports and remediated the specific Trusted URLs bypass used in the research. Zenity said…

Excerpt shown under the publisher's RSS terms · 61 words · full text and figures are on the publisher's site

Read at Cybersecurity News ↗Ask about this article

Related coverage

AI Safety

How ChatGPT's new Lockdown mode protects you from data theft (and what else it does)

ZDNet · Jun 8, 2026

AI Safety

Friendly Fire: Hijacking Defensive Cyber AI Agents for Remote Code Execution

AI Now Institute · Jul 8, 2026

Weekly Cybersecurity Newsletter Bulletin – Cisco and Android 0-Day, BragJack Attack, Claude Opus 5 Used to Hack OpenAI, and 20+ Stories

Cybersecurity News · Sep 21, 2026

AI Safety

Zero-Shot Embedding Drift Detection: A Lightweight Defense Against Prompt Injections in LLMs

arXiv · Jun 8, 2026