Ramp's AI-powered Sheets tool exposed sensitive financial information through unintended data exfiltration. The vulnerability allowed unauthorized access to confidential business records.
Security researchers identified a flaw in Ramp's Sheets AI feature that inadvertently leaked financial data from user accounts. The issue stemmed from inadequate data isolation in the AI processing pipeline, enabling information to be extracted beyond intended scope.
Ramp, a corporate spend management platform, integrates AI capabilities to analyze and organize financial spreadsheets. The vulnerability affected users' sensitive financial records, including transaction data and account details.
The discovery was documented in a detailed analysis by PromptArmor, a security research firm specializing in AI vulnerabilities. The report outlines how the exfiltration occurred and the potential exposure window.
Ramp has not yet issued a public statement regarding the incident or confirmation of when the vulnerability was discovered and patched. Users of the platform's AI features should review their account activity and financial data access logs.
The incident highlights ongoing security concerns surrounding AI integration in fintech platforms and the importance of rigorous data isolation protocols.
A Cambridge study reveals that terrorist organizations including Boko Haram and ISIS are using ChatGPT, Claude, and Gemini to plan attacks and develop weapons. Safety filters designed to prevent such misuse have repeatedly failed.
The Australian Cyber Security Centre has issued an alert about coordinated exploitation of vulnerable content management systems and plugins worldwide. The campaign targets organizations using outdated or unpatched CMS software.
Artificial intelligence discovered a critical security vulnerability in Linux kernel code that human developers overlooked for over a decade. The bug could allow unauthorized root access to systems.
Researchers have demonstrated a new attack called 'Ghostcommit' that hides prompt injections in PNG files to fool AI code reviewers and agents into exposing repository secrets.