‘SalesBleed’ Flaws in Salesforce Agentforce Enabled Zero-Click Data Exfiltration
Salesforce Agentforce had 'SalesBleed' flaws enabling zero-click data exfiltration and phishing.
Three 'SalesBleed' vulnerabilities in Salesforce Agentforce allowed attackers to hijack trusted agents for data exfiltration and phishing attacks. Exploitable via Web-to-Lead forms, two flaws enabled zero-click data theft by bypassing Trusted URLs, while the third weaponized the Agentforce-Slack integration for phishing. Salesforce has since patched these issues.
AI summary of the linked source. Not reviewed by an editor. How this works
Run Salesforce?
Get an email when a new story names Salesforce, usually within the hour.
Free. Your list stays private and never appears in a subject line. One click stops it. How Vendor Watch worksPrivacy
The lesson from this incident
Full lessonThree vulnerabilities in Salesforce Agentforce—collectively dubbed 'SalesBleed'—allowed attackers to hijack trusted AI agents to exfiltrate sensitive data and launch phishing campaigns without requiring any user interaction. Two flaws exploited Web-to-Lead forms to bypass Trusted URL controls, enabling zero-click data theft, while a third weaponized the Agentforce-Slack integration as a phishing vector. This incident highlights the expanding attack surface introduced by AI-driven automation platforms and third-party integrations, which can inherit or amplify trust relationships in dangerous ways. Organizations must treat AI agent configurations and integration points as first-class security assets requiring rigorous review and timely patching.
What helps against this
How we list vendors- Data loss prevention (DLP)
Sensitive data cannot quietly leave by email, upload, USB stick or AI tool.
- AI application security
Prompt injection, data leaks through AI tools and unapproved AI use get caught.
- Email security
Phishing, malware and invoice fraud are filtered out before they reach inboxes.