MIT Study Finds AI Agents Face Major Security Gaps in 2025
A landmark MIT-led study of 30 AI agents reveals critical gaps in transparency, security testing, and user control across the agentic AI industry.
Microsoft Copilot Bug Exposed Confidential Emails in Outlook
A Microsoft 365 Copilot bug let AI summarize confidential emails in Sent and Drafts folders, bypassing DLP security policies.
Cisco unveils Splunk AI security editions
Cisco launches Splunk AI security editions with faster response, unified workflows and global availability.