Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

AI agent security solutions fall into two categories. Some use AI agents to perform security work, such as red teaming, pentesting, SOC investigation, threat hunting, and risk analysis. Others protect AI agents, copilots, MCP servers, and agentic workflows from vulnerabilities such as over-permissioning, prompt injection, unsafe tool use, data exposure, and unauthorized actions.

Veil#Drop is a sophisticated multi-stage malware delivery framework that combines social engineering, compromised websites, malicious JavaScript launchers, PowerShell download cradles, and trusted cloud-hosted infrastructure to deploy PureLog Stealer entirely in memory. The infection chain begins with a deceptively named JavaScript file masquerading as a document (e.g., transcript.pdf.js), which executes through Windows Script Host and launches PowerShell with execution policy bypasses enabled.

A lot of security programs still treat data exfiltration as a downstream consequence of compromise. That framing is too narrow. The global average cost of a breach reached $4.44 million in 2025 according to Varonis's summary of 2025 data breach statistics, and that cost lands on operations, legal, compliance, and executive credibility, not just the SOC.

AI today has moved beyond experimentation. In the modern age, enterprises are embedding AI across various aspects of their businesses, including customer support, document processing, software development, healthcare, financial services, and decision-making workflows. According to a recent McKinsey report, 88% of businesses use AI in at least one business function. This reflects how AI is now becoming the center of several enterprise operations.

A new report from SpyCloud has found that phishing attacks have exposed employee data at 86% of Fortune 100 companies over the past 12 months, with the technology, airline and automotive sectors being hit the hardest. The researchers also found that 78% of organizations experienced an increase in phishing volume over the past year. Additionally, 84% of respondents named AI-assisted phishing as their top concern, followed by business email compromise (BEC) attacks.

I saw a venn diagram on social media. One circle is Shadow IT, one circle is Shadow AI, a substantial overlap, and the implicit message is that they are effectively the same challenge. They aren’t and that the assumption can lead to many problems. Looking back, shadow IT was like watching a crash in slow-motion. Employees using technology IT hadn't sanctioned. Personal Dropbox accounts. Unofficial Slack workspaces.

The 2026 NASCIO-Deloitte Cybersecurity Study reveals a stark reality for CISOs in state governments: while cyber threats are growing in both sophistication and volume, the resources available to combat them are failing to keep pace. As foreign adversaries and cybercriminals weaponize AI to probe for vulnerabilities, state CISOs find themselves at a critical juncture, navigating expanding responsibilities amidst tightening budgets.

The LogRhythm SIEM July 2026 release adds new investigation workflow features, expands automation for administration and archiving, and broadens telemetry coverage across cloud, identity, collaboration, endpoint, and email environments. Organizations running on-premises and hybrid environments often need tight control over data to meet sovereignty and operational requirements.

Low and slow attacks look normal because they are intentionally distributed into small, permissible actions that avoid detection thresholds. Each step appears legitimate on its own, which prevents detection systems from recognizing the overall progression. The issue is not that security teams lack telemetry. The issue is that traditional detection often evaluates activity in fragments. When each action stays below a rule or threshold, the broader pattern can remain invisible.

Cato AI Labs has discovered two critical remote code execution (RCE) vulnerabilities in Cursor IDE, the popular development environment which, according to Cursor, is used by over half of the Fortune 500. Both RCE vulnerabilities, which we refer to as “DuneSlide,” achieved a 9.8 CVSS score, and involve breaking out of the IDE’s sandbox environment and were assigned CVE IDs CVE-2026-50548 and CVE-2026-50549.