Skip to content

Sam's News β€” security β€” 2026-09-04

Security

8 IDScan Data Breach Exposes 153 Million Driver's Licenses

Identity verification company IDScan faces multiple lawsuits after hackers allegedly breached the system and stole 153 million driver's licenses.

Sources: BleepingComputer RSS

8 HPE Patches Critical RCE Vulnerabilities in AOS-CX Switches

HPE patched 34 CVEs in Aruba AOS-CX, including nearly two dozen critical remote code execution flaws with CVSS 9.8 that allow unauthenticated attackers to execute commands via malformed packets. Additional updates address 22 high-severity and 11 medium-severity vulnerabilities across multiple AOS-CX versions.

  • ~24 critical RCE vulnerabilities, CVSS score 9.8, tracked as CVE-2026-73749
  • Unauthenticated attacker can achieve RCE with elevated privileges via crafted packets
  • 34 total CVEs patched; also includes 22 high-severity and 11 medium-severity flaws
  • Affects AOS-CX versions 10.18.1002, 10.17.1030, 10.16.1060, 10.13.1190, 10.10.1181
  • No known exploitation in the wild; majority discovered internally by HPE

Sources: SecurityWeek AI Web Searched

8 Citrix NetScaler critical authentication bypass exploited in active attacks

Attackers are actively exploiting CVE-2026-19490, a critical authentication bypass vulnerability in Citrix NetScaler.

Sources: BleepingComputer RSS

7.5 PostgreSQL hit by 12-year-old vulnerability enabling server takeover

PostgreSQL is affected by a decade-old security vulnerability that permits complete server compromise through logical decoding.

Sources: securityaffairs.com RSS, The Hacker News RSS Update to: Decade-old PostgreSQL vulnerability enables backup account backdoor access

7 Virtual Machines Insufficient to Contain Advanced AI Agents

Security researchers, including Bruce Schneier, found that standard virtual machines lack sufficient containment for advanced AI agents like GPT 5.6-Cyber, as even benign features such as display functionality introduce exploitable vulnerabilities. The findings suggest fundamental reassessment of sandboxing quality in software stacks used by cyber-enabled AI agents.

  • Standard VMs insufficient to contain GPT 5.6-Cyber and similar advanced agents
  • Display functionality and other benign VM features introduce exploitable attack surface
  • Cyber-enabled AI agents' software stack requires fundamental sandboxing reassessment

Sources: Schneier on Security AI Web Searched

7 OpenAI Pledges $1 Billion Daybreak Initiative for Critical Infrastructure AI Security

OpenAI announced the Daybreak initiative, pledging $1 billion to provide subsidized AI capabilities, training, and technical assistance to critical infrastructure defenders.

Sources: SecurityWeek RSS

7 CISA warns of old ownCloud vulnerability after Philippine nuclear data theft report

CISA flagged an old ownCloud vulnerability following reports of a data theft incident involving Philippine nuclear information.

Sources: TechRepublic RSS

7 High-volume phishing campaign uses invisible Unicode to bypass email filters

Microsoft reported a large-scale phishing campaign exploiting invisible Unicode tag characters to evade email security filters and deliver financial fraud lures.

  • High-volume phishing campaign started February 9, 2026
  • Uses invisible Unicode tag characters (U+E0000–U+E007F) to split keywords like 'funding'
  • Technique adapted from AI prompt injection research
  • Microsoft Defender detections elevated for approximately three months
  • Most flagged messages caught by layered protections rather than single Unicode signals

Sources: The Hacker News RSS Update to: Unicode smuggling technique adapted for email phishing evasion

7 SonicWall warns of actively exploited vulnerabilities in SMA1000 appliances

SonicWall issued warnings about actively exploited security vulnerabilities affecting its SMA1000 remote access appliances.

Sources: The HIPAA Journal RSS Update to: SonicWall SMA 1000 Critical Zero-Day Enables Unauthenticated Remote Compromise

6 Astra launches, CISA cuts programs, McKesson breach in security briefing

Recent developments include OpenAI's Astra launch, CISA program reductions, and a confirmed McKesson data breach.

Sources: CISO Series RSS