Security News
Analysis of AI security incidents, vulnerabilities, and industry news.
9 articles
Salesloft-Drift Breach: How a Single AI Chatbot Exposed 700+ Enterprise Customers
The August 2025 Salesloft-Drift breach became one of the largest SaaS supply-chain attacks in history, affecting Cloudflare, Palo Alto Networks, Zscaler, and hundreds more.
EchoLeak: The Zero-Click Microsoft 365 Copilot Vulnerability That Changed AI Security
A critical vulnerability in Microsoft 365 Copilot enabled attackers to steal sensitive corporate data without any user interaction. What it means for enterprise AI security.
AI Chatbots Top ECRI's 2026 Health Technology Hazards List
Healthcare's leading patient safety organization warns that AI chatbots pose the greatest health technology risk in 2026. Here's what healthcare AI teams need to know.
Slack AI Vulnerable to Prompt Injection: Private Channel Data at Risk
Security researchers discovered Slack AI could be tricked into leaking data from private channels via prompt injection. Analysis of the vulnerability and defense strategies.
Klarna's AI Customer Service Experiment: Why Replacing 700 Humans Backfired
Swedish fintech giant Klarna fired 700 support agents to go AI-first. Customer satisfaction plummeted, complaints surged, and they had to rehire humans. What went wrong.
120+ Legal AI Hallucination Cases: Why Courts Are Sanctioning Lawyers for Fake Citations
Over 120 cases of AI-generated legal hallucinations have been identified since 2023, with sanctions reaching $31,100. Stanford research shows even premium legal AI tools hallucinate 17-34% of the time.
AI-Powered Fraud Surge: $12.5 Billion in Losses and the Rise of FraudGPT
Consumers lost $12.5 billion to fraud in 2025, with AI tools like FraudGPT enabling scams at unprecedented scale. What fintech and security teams need to know.
GitHub Copilot Security: 39% of AI-Generated Code Contains Vulnerabilities
Research reveals that AI code assistants introduce security vulnerabilities at alarming rates, and a new supply chain attack vector weaponizes code agents against developers.
Volkswagen Cariad: $7.5 Billion Lessons from Automotive's Biggest AI Software Failure
Volkswagen's ambitious AI-driven operating system project lost $7.5 billion over three years. What went wrong and what it teaches about AI system testing and deployment.