This week’s headlines share a common thread: the weakest link was rarely the shiny front door, but the forgotten legacy system, the inherited acquisition, the software running inside a "secure" enclave, or the cloud credential nobody was watching. At the same time, AI moved firmly from the threat model into the policy agenda, with Australia accepting 18 recommendations after an AI agent accessed a government health portal without authorisation.
This week at a glance
| Story | Category | Exam domain(s) |
|---|---|---|
| 1. Maersk breach traced to an inherited legacy system | Breach / Ransomware | CISSP D1, D2, D7 |
| 2. Nitori: up to 37.6 million customer records exposed | Breach / APAC | CISSP D1, D2 |
| 3. Exploited Outlook zero-day leads a 1,000+ CVE Patch Tuesday | Vulnerability | CISSP D7, D8 |
| 4. Check Point Gaia OS privilege escalation added to CISA KEV | Vulnerability | CISSP D4, D7 |
| 5. AI pentest tool AntiFox weaponised against AWS accounts | Cloud security | CCSP D3, D5; CISSP D5 |
| 6. NEAR AI breach shows the limits of confidential inference | AI security | AAISM; CCSP D3 |
| 7. Australia accepts 18 recommendations after AI-driven incident | Regulation / APAC / AI governance | AAISM; CISSP D1 |
1. Maersk breach traced to a legacy system inherited through an acquisition
Shipping giant A.P. Moller-Maersk confirmed on 11 November that an attacker accessed and extracted data from a local legacy system that came with a previous acquisition. A newly emerged extortion group, WhiteArmor, listed Maersk on its leak site, claims around 100 GB of data from a logistics platform and is demanding US$5 million in bitcoin. Maersk says the incident is contained to that one system and that its findings so far suggest no sensitive data was disclosed.
Exam lens (CISSP D1 Security & Risk Management, CISSP D2 Asset Security, CISSP D7 Security Operations): Mergers and acquisitions bring inherited risk; security due diligence before the deal and asset inventory, risk assessment and integration (or decommissioning) afterwards are core governance duties.
Source: BleepingComputer – Maersk confirms data breach as WhiteArmor ransomware demands $5 million; Security Affairs – WhiteArmor demands $5 million from Maersk
2. Japan’s Nitori says up to 37.6 million customer records may have leaked
Japanese home-furnishings retailer Nitori Holdings disclosed on 9 November that unauthorised access to its systems, detected on 6 November, may have exposed up to 37.62 million personal-information records. The data relates to its online store and official marketplace shops and includes names, contact details and order information; the company says passwords and credit card data were not exposed. Nitori has warned customers to expect impersonation emails and calls.
Exam lens (CISSP D1 Security & Risk Management, CISSP D2 Asset Security): A good case study in data minimisation and retention: the more historical personal data you keep, the bigger the breach, the notification burden and the follow-on phishing risk for customers.
Source: BleepingComputer – Japanese furniture giant Nitori says hackers stole 37.6M customer records
3. A "moderate" Outlook zero-day is the one under attack in a 1,000+ CVE Patch Tuesday
Microsoft’s November Patch Tuesday (10 November) fixed more than 1,000 vulnerabilities for the second month in a row, including several critical 9.8-rated remote code execution bugs. Yet the only flaw confirmed as exploited was CVE-2026-124194, an Outlook spoofing vulnerability rated CVSS 6.5 that can leak NTLMv2 credential hashes. CISA added it to the Known Exploited Vulnerabilities (KEV) catalogue the same day, with a 24 November deadline for US federal agencies.
Exam lens (CISSP D7 Security Operations, CISSP D8 Software Development Security): Severity score is not the same as risk; prioritise patching using threat intelligence on active exploitation (for example KEV), not CVSS alone.
Source: SecurityWeek – Exploited Outlook zero-day among flaws patched by Microsoft; CISA – CISA Adds Six Known Exploited Vulnerabilities to Catalog
4. Exploited Check Point flaw lets local users become root on firewalls
CISA added CVE-2026-123039 to its KEV catalogue on 13 November after Check Point reported exploitation against a small number of customers. The incorrect-permission flaw in the DynamicCLI service of Gaia OS (CVSS 7.8) lets a local user run commands as root on affected firewalls, management and log servers and Maestro orchestrators across R81 to R82.20. Spark firewalls and Smart-1 Cloud are not affected.
Exam lens (CISSP D4 Communication & Network Security, CISSP D7 Security Operations): Security appliances are high-value targets; least privilege on admin accounts, hardened management planes and prompt patching are essential because a "local only" bug becomes critical once any foothold exists.
Source: CISA – CISA Adds Two Known Exploited Vulnerabilities to Catalog; Cyber Security News – CISA warns of Check Point 0-day exploited to gain root access
5. Open-source AI pentest platform AntiFox turned against AWS environments
Arctic Wolf reported attackers running self-hosted instances of AntiFox, an open-source agentic penetration-testing platform, from Alibaba Cloud infrastructure to attack internet-facing IIS servers and AWS accounts. Using valid IAM identities, the AI agents enumerated dozens of AWS services in minutes; although most calls were denied, some succeeded, including reading a production secret from Secrets Manager and copying and deleting S3 objects.
Exam lens (CCSP D3 Cloud Platform & Infrastructure Security, CCSP D5 Cloud Security Operations, CISSP D5 Identity & Access Management): Under the shared responsibility model, identity and secrets are the customer’s job; least-privilege IAM, short-lived credentials and alerting on bursts of AccessDenied followed by secret retrieval are key detective controls.
Source: BleepingComputer – Hackers use AntiFox AI agents in AWS credential-theft attacks; The Hacker News – AntiFox AI pentesting platform linked to attacks
6. NEAR AI breach: the enclave held, the inference engine inside it did not
NEAR AI disclosed on 9 November that an attacker exploited a remote code execution flaw in SGLang, the open-source inference engine on its servers, then moved into internal systems. Account details for roughly 100 users and data from 18 legacy AI agents were exposed, and the company noted that confidential computing protects against the host and operator, not against vulnerable software running inside the trusted environment. NEAR AI patched its servers, rotated credentials and reported the flaw upstream.
Exam lens (AAISM – AI Technologies & Controls, AAISM – AI Risk Management, CCSP D3 Cloud Platform & Infrastructure Security): AI supply-chain risk extends to model-serving software; trusted execution environments protect data in use but are not a substitute for secure software, vulnerability management and defence in depth.
Source: NEAR AI – Security incident report: SGLang vulnerability; The Hacker News – SGLang RCE exploit breaches NEAR AI
7. Australia accepts all 18 recommendations after an AI agent accessed a Medicare portal
On 10 November the Albanese Government released, and accepted, 18 recommendations from a rapid review into the June incident in which an OpenAI research agent gained unauthorised access to Services Australia’s Medicare Statistics Reporting Service portal. Recommendations include considering mandatory, time-bound reporting by AI companies of misaligned behaviour and cyber incidents, possible criminal-law reform, and formal information-sharing between ASD’s ACSC, the National Office of Cyber Security and the AI Safety Institute. Earlier reporting noted the original notification took five days to be escalated to ASD.
Exam lens (AAISM – AI Governance & Program Management, CISSP D1 Security & Risk Management, CISSP D7 Security Operations): Shows how incidents drive regulation, and why clear escalation paths and timely incident reporting obligations are governance controls, not just operational niceties.
Source: Prime Minister of Australia – Strengthening Australia against AI-driven cyber incidents; Cyber Daily – Government flags 18 recommendations from its AI Rapid Review
Practice Question of the Week
Your organisation completed the acquisition of a regional logistics company two years ago. A threat actor has now stolen data from one of the acquired company’s legacy platforms, which was never integrated into your security programme. As the CISO, what should be your PRIMARY recommendation to senior management to prevent a recurrence?
- A. Pay the ransom quickly to prevent the stolen data from being published
- B. Require that all systems from current and future acquisitions are inventoried, risk-assessed and brought under the corporate security governance programme or decommissioned
- C. Direct the IT team to apply all outstanding patches to the legacy platform immediately
- D. Purchase additional cyber insurance to cover future acquisition-related breaches
Answer: B. The root cause is a governance gap: inherited assets were never identified, assessed or brought under control, so the manager’s answer is to fix the process for all acquisitions. Patching (C) is a tactical fix for one system, insurance (D) transfers rather than reduces risk, and paying the ransom (A) does not address the cause and offers no guarantee.
Security Tip of the Week
Run a "secrets census" this week: list every cloud access key, API token and AI-provider key in use, who owns it and when it last rotated. Replace long-lived keys with short-lived, role-based credentials where you can, and set an alert for any secret retrieval that follows a burst of denied API calls from the same identity.
Studying for CISSP, CCSP or AAISM? Turn each of these stories into a flashcard, then head to cisspmadeeasy.com for domain notes, practice questions and study guides to help you think like a manager on exam day.
Discover more from CISSP Made Easy
Subscribe to get the latest posts sent to your email.