Cloud Platforms Tighten Defenses as CISA Flags Critical ICS Risks

Cloud Armor: Hierarchical Policies, Extended WAF and NTI

🛡️ Cloud Armor introduces hierarchical security policies and organization-scoped address groups to simplify centralized policy management across organization, folder, and project levels. The release also includes GA support for JA4 network fingerprinting and ASN/NTI controls for Media CDN, while an enhanced WAF request-body inspection (preview) expands inspection from 8 KB to 64 KB. These updates are designed to strengthen threat protection and reduce operational complexity for hybrid and multicloud deployments.

Microsoft Security Store Unites Partners and Innovation

🔐 Microsoft Security Store, released to public preview on September 30, 2025, is a unified, AI-powered marketplace that lets organizations discover, buy, and deploy vetted security solutions and AI agents. Catalog items — organized by frameworks like NIST and by integration with products such as Microsoft Defender, Sentinel, Entra, and Purview — address threat protection, identity, compliance, and cloud security. Built on the Microsoft Marketplace, it provides unified billing, MACC eligibility, and guided automated provisioning to streamline deployments.

Raisecomm RAX701-GC SSH Authentication Bypass Vulnerability

🔒 A critical authentication bypass in Raisecomm RAX701-GC devices permits SSH sessions without completing user authentication, potentially granting unauthenticated root shell access. The flaw is tracked as CVE-2025-11534 with a CVSS v3.1 score of 9.8 and CVSS v4 score of 9.3, exploitable remotely with low attack complexity. Affected firmware versions include 5.5.27_20190111, 5.5.13_20180720, and 5.5.36_20190709. CISA recommends isolating affected devices from the internet, placing control networks behind firewalls, and using secure remote access methods such as updated VPNs while contacting vendor support.

AWS Nitro Enclaves Now Available in All Regions Worldwide

🔒 AWS has made Nitro Enclaves available in every AWS Region, expanding regional support to include new locations across Asia Pacific, Europe, the Middle East, and North America. Nitro Enclaves enables customers to create isolated compute environments inside EC2 instances to protect and process sensitive data and reduce attack surface. There is no additional charge beyond the EC2 and associated service usage.

Amazon Nova adds customizable content moderation settings

🔒 Amazon announced that Amazon Nova models now support customizable content moderation settings for approved business use cases that require processing or generating sensitive content. Organizations can adjust controls across four domains—safety, sensitive content, fairness, and security—while Amazon enforces essential, non-configurable safeguards to protect children and preserve privacy. Customization is available for Amazon Nova Lite and Amazon Nova Pro in the US East (N. Virginia) region; customers should contact their AWS Account Manager to confirm eligibility.

The Signals Loop: Fine-tuning for AI Apps and Agents

🔁 Microsoft positions the signals loop — continuous capture of user interactions and telemetry with systematic fine‑tuning — as essential for building adaptive, reliable AI apps and agents. The post explains that simple RAG and prompting approaches often lack the accuracy and engagement needed for complex use cases, and that continuous learning drives sustained improvements. It highlights Dragon Copilot and GitHub Copilot as examples where telemetry‑driven fine‑tuning yielded substantial performance and experience gains, and presents Azure AI Foundry as a unified platform to operationalize these feedback loops at scale.

CloudWatch Database Insights: On-Demand Analysis for RDS

📊 Amazon CloudWatch Database Insights now offers on-demand analysis for RDS for SQL Server, extending automated diagnostics to that engine. The feature uses machine learning to compare a selected time period against baseline performance, surface anomalies, and provide tailored remediation advice. Administrators can enable this in Advanced mode via the RDS console, APIs, SDKs, or CloudFormation to reduce mean-time-to-diagnosis from hours to minutes.

Amazon EC2 U7i-6TB High Memory Instances in London

🚀 AWS has launched the U7i-6tb High Memory instance in the Europe (London) Region, offering 6TB of DDR5 memory and 448 vCPUs for large in-memory workloads. Powered by custom fourth-generation Intel Xeon Scalable (Sapphire Rapids) processors, these 7th-generation instances support up to 100 Gbps for EBS and network and include ENA Express for lower latency. They are aimed at mission-critical in-memory databases such as SAP HANA, Oracle, and SQL Server, enabling higher transaction throughput and faster data loading and backups.

Deploying AWS Secrets Manager Agent as an EKS Sidecar

🔒 This post demonstrates deploying the AWS Secrets Manager Agent as a sidecar container in Amazon EKS to provide a language-agnostic local HTTP interface (localhost:2773) for secrets retrieval. The agent pulls and caches secret values, reducing direct API calls to Secrets Manager and improving application availability. It enforces SSRF protection via a generated token at /var/run/awssmatoken and implements ML‑KEM post‑quantum key exchange by default. Authentication uses Amazon EKS Pod Identity and IAM permissions (secretsmanager:GetSecretValue and secretsmanager:DescribeSecret), and the post includes build, containerization, and deployment steps.

CloudEdge Online Cameras/App—MQTT Wildcard Credential Risk

🔒 The CloudEdge mobile app (v4.4.2) and associated online cameras contain a credential exposure flaw assigned CVE-2025-11757 that stems from improper MQTT topic handling (CWE-155). Unsanitized topic input allows an attacker to use MQTT wildcards to subscribe to other users' messages and extract credentials and key material, enabling remote access to live feeds and camera controls. CISA calculated a CVSS v4 base score of 8.7 and highlights low attack complexity and remote exploitability. Users are advised to minimize network exposure, isolate devices behind firewalls, employ secure remote access methods such as VPNs with caution, and contact Meari Technologies support at support@mearitek.com.

Critical WatchGuard Fireware OS RCE via IKEv2 VPN Exploit

🔴 A critical out-of-bounds write vulnerability (CVE-2025-9242) in WatchGuard Fireware OS could allow remote code execution via IKEv2 mobile VPN and Branch Office VPN when configured with dynamic gateway peers. Affected releases include Fireware OS 11.10.2 through 11.12.4_Update1, 12.0 through 12.11.3 and 2025.1, and WatchGuard warns devices previously configured with these peers may remain vulnerable. Shadowserver estimates over 71,000 potentially exposed devices; WatchGuard and the US NVD have published advisories and guidance, and a temporary workaround plus narrower BOVPN access policies are recommended if immediate upgrades are not possible.

CISA Confirms Exploitation of Oracle E-Business SSRF Flaw

🔒 CISA has confirmed active exploitation of CVE-2025-61884, an unauthenticated SSRF in the Oracle Configurator runtime, and added it to its Known Exploited Vulnerabilities catalog. Federal agencies are required to patch the issue by November 10, 2025. Oracle released a fix on October 11 rated 7.5 and BleepingComputer says the update blocks a leaked exploit tied to ShinyHunters and related extortion activity.

UK Contractor Breach Exposes Sensitive RAF and Navy Sites

🔒 A ransomware attack on contractor Dodd Group reportedly allowed Russian-linked attackers to exfiltrate hundreds of sensitive Ministry of Defence documents, including details on RAF Lakenheath, RAF Portreath and RAF Predannack. The company confirmed an incident and said it contained access, while the MoD suspects the Lynx group is behind the intrusion. Leaked files published on the dark web allegedly include site plans and personnel data, and the case is now under investigation amid a wider rise in UK cyber incidents.

Microsoft October 2025 Patch Causes Enterprise Failures

🚨 The October 2025 Windows security update KB5066835, intended to move cryptography from CSP to KSP, is causing widespread enterprise disruption. Affected platforms — including Windows 10 (22H2), Windows 11 (23H2–25H2) and several Windows Server releases — report smartcard and certificate failures, USB mouse/keyboard loss in WinRE, IIS ERR_CONNECTION_RESET and WUSA installation errors. Microsoft published a registry workaround (DisableCapiOverrideForRSA=0) and an out‑of‑band update (KB5070773) for some issues, but urges caution and recommends thorough testing before broad deployment.

CISA Releases 10 ICS Advisories Covering Multiple Vendors

🔔 CISA released 10 Industrial Control Systems (ICS) advisories providing technical details about vulnerabilities, impacts, and mitigations affecting multiple vendors. Notable entries include Rockwell Automation products (1783-NATR, Compact GuardLogix 5370), Siemens devices (SIMATIC S7-1200, RUGGEDCOM ROS), Schneider Electric Modicon controllers and HMI software, plus camera and networking products. Administrators should review each advisory and apply recommended mitigations promptly.

Siemens SIMATIC S7-1200 Vulnerabilities and Patches Updates

⚠️ Siemens has published an advisory for SIMATIC S7-1200 CPU V1/V2 devices describing two high-severity vulnerabilities: an Improper Input Validation flaw (CVE-2011-20001) that can force a controller into a stop/defect state via malformed HTTP traffic, and an Authentication Bypass by Capture-Replay (CVE-2011-20002) that allows replay of engineering commands. CVSS v4 scores are high (up to 8.7); Siemens recommends updating firmware (V2.0.3/V2.0.2) and disabling the web server where possible, while CISA advises network segmentation, firewalling, and avoiding direct Internet exposure.

Rockwell Automation 1783-NATR: Critical Remote Flaws

⚠️ Rockwell Automation's 1783-NATR network adapter contains multiple high-severity vulnerabilities, including missing authentication for critical functions, stored XSS, and CSRF. CISA assigns CVSS v4 9.9 for the most severe issue and warns these flaws can be exploited remotely with low complexity to cause denial-of-service, data modification, or credential compromise. Rockwell Automation recommends upgrading to 1.007 or later; CISA advises minimizing network exposure and isolating control networks.

Russian Star Blizzard shifts to 'Robot' malware families

🔐 The Russian state-backed Star Blizzard group (aka ColdRiver/UNC4057) has shifted to modular, evolving malware families — NOROBOT, YESROBOT, and MAYBEROBOT — delivered through deceptive ClickFix pages that coerce victims into executing a fake "I am not a robot" CAPTCHA. NOROBOT is a malicious DLL executed via rundll32 that establishes persistence through registry changes and scheduled tasks, stages components (including a Windows Python 3.8 install), and, after iteration, primarily delivers a PowerShell backdoor. Google Threat Intelligence Group and Zscaler observed the transition from May through September and reported that ColdRiver abandoned the previously exposed LostKeys tooling shortly after disclosure. GTIG has published IoCs and YARA rules to help defenders detect these campaigns.

TP-Link Omada Gateways Vulnerable to Critical RCE Flaw

⚠️ TP-Link has disclosed two command injection vulnerabilities affecting Omada gateway devices that allow execution of arbitrary OS commands. One issue, CVE-2025-6542 (CVSS 9.3), can be exploited remotely without authentication; the other, CVE-2025-6541 (CVSS 8.6), requires access to the web management interface. Thirteen models are listed as impacted and TP-Link has released firmware updates to address the flaws; administrators are urged to apply patches and verify configurations after upgrading.

Cursor, Windsurf IDEs Exposed to 94+ Chromium Flaws

⚠️ The latest releases of Cursor and Windsurf IDEs embed outdated Chromium and V8 engines that contain at least 94 known, patched vulnerabilities. Ox Security researchers demonstrated a proof‑of‑concept exploiting CVE-2025-7656 (a Maglev JIT integer overflow) to crash Cursor, and warn that similar flaws could enable denial‑of‑service or arbitrary code execution in real attacks. Attack vectors include deeplinks, malicious extensions, poisoned README previews or documentation; the two IDEs together serve an estimated 1.8 million developers. Cursor dismissed the DoS finding as out of scope and Windsurf did not respond to inquiries.

Siemens RUGGEDCOM TLS and Access Control Vulnerabilities

🔒 Siemens published an advisory (republished by CISA) for multiple vulnerabilities affecting RUGGEDCOM ROS devices, including CVE-2023-52236 and several CVE-2025-4122x issues. The flaws involve risky cryptographic algorithms, improper TLS handshake handling that can cause DoS, and an access-control enforcement failure that persists until reboot. Siemens has released updates (V5.10.0+) for many models and recommends restricting management ports, disabling web/SSH services if unused, and configuring GCM ciphers where applicable. CISA reiterates standard ICS guidance to minimize network exposure and isolate control networks.

Rockwell Compact GuardLogix 5370 Uncaught Exception

⚠️ Rockwell Automation has disclosed an uncaught exception vulnerability in Compact GuardLogix 5370 controllers that can be triggered by a crafted CIP unconnected explicit message and may cause a non‑recoverable fault resulting in denial-of-service. The issue is tracked as CVE-2025-9124 and carries a CVSS v4 base score of 8.7, indicating remote exploitability with low complexity. Rockwell recommends upgrading affected devices to firmware 30.14 or later; organizations unable to upgrade should follow vendor security best practices and apply network isolation measures.

Critical MinKNOW Vulnerabilities Allow Remote Access and DoS

⚠️ Oxford Nanopore Technologies MinKNOW sequencing software contains multiple remotely exploitable vulnerabilities (highest CVSS v4 8.3) that can permit unauthorized access, data manipulation, and denial-of-service on affected devices. Attackers can discover devices via network scanning, exploit authentication that trusts host IPs, and reuse tokens stored in world-readable temporary files to gain persistent access or redirect sequencing output. Oxford Nanopore advises upgrading to versions later than 24.11; CISA recommends minimizing network exposure, isolating control networks, and using secure remote access methods while applying other mitigations.

Coldriver Deploys New 'NoRobot' Malware Suite, 2025

🛡️ Google Threat Intelligence Group (GTIG) has observed the Russian-linked Coldriver group deploying a new, staged malware ecosystem tracked as NoRobot, YesRobot and MaybeRobot. GTIG's October 20, 2025 report shows the campaign replaces the previously disclosed LostKeys strain and begins with a 'ClickFix-style' ColdCopy phishing lure that tricks victims into running a malicious DLL via rundll32.exe. NoRobot functions as a downloader using split-key cryptography and staged payloads; operators briefly used a Python-based backdoor (YesRobot) before switching to a more flexible PowerShell backdoor (MaybeRobot) to reduce detection.

CrowdStrike Launches AI-Driven Falcon UX in Preview

🔍 At Fal.Con 2025, CrowdStrike introduced a dynamic, persona-aware user experience for Falcon Cloud Security and Falcon Exposure Management, now available in public preview. Built on CrowdStrike Enterprise Graph and Charlotte AI, the console unifies hybrid and multi-cloud asset and risk visibility into customizable workspaces. It offers AI-assisted dashboard creation and executive-ready reporting to accelerate investigations and remediation without switching tools.

Dataplex Supports Column-Level Lineage for BigQuery

🔍 Dataplex Universal Catalog now captures column-level lineage for BigQuery, extending object-level tracing to granular column transformations at no extra cost. The update provides interactive visual lineage graphs so users can inspect upstream and downstream flows for individual columns, trace origins, and assess downstream impact of modifications. This granularity helps validate authoritative sources for AI/ML features, enforce column-level governance, and improve compliance. It also surfaces freshness and usage metadata to support context-aware agents.

PolarEdge Botnet Targets Cisco, ASUS, QNAP Routers

🔐 Cybersecurity researchers have detailed PolarEdge, a TLS-based ELF implant used to conscript Cisco, ASUS, QNAP and Synology routers into a botnet. The backdoor implements an mbedTLS v2.8.0 server with a custom binary protocol, supports a connect-back and interactive debug mode, and stores its obfuscated configuration in the final 512 bytes of the ELF. Operators use anti-analysis techniques, process masquerading and file-moving/deletion routines; a forked watchdog can relaunch the payload if the parent process disappears.

Vidar Stealer 2.0 Rewritten in C with Multi-Threading

🛡️ Vidar Stealer 2.0 was released with a complete rewrite in C, multi-threaded data theft and stronger evasion, prompting warnings from security researchers about likely increased campaigns. The update reduces dependencies and footprint while spawning parallel worker threads to accelerate harvesting of browser, wallet, cloud and app credentials. It introduces extensive anti-analysis checks and a polymorphic builder to frustrate static detection. Notably, the malware injects into running browser processes to extract encryption keys from memory and bypass Chrome's App-Bound protections.

John Bolton Charged Over Classified Emails Leak After Hack

🔒Former national security adviser John Bolton has been charged with mishandling classified information after prosecutors say he retained and transmitted sensitive documents via a personal AOL account that was later accessed by suspected Iranian hackers. The intruders allegedly downloaded the materials and sent extortion messages to Bolton. The case highlights questions about password strength, the use of two-step verification, and the risks of sending unencrypted, sensitive information to family members. Bolton has pleaded not guilty.

Reducing Abuse of Microsoft 365 Exchange Online Direct Send

🛡️ Cisco Talos warns that Microsoft 365 Exchange Online’s Direct Send feature, intended for legacy devices and line‑of‑business appliances, is being abused to bypass standard authentication and content inspection. Attackers are leveraging these unauthenticated SMTP flows in phishing and BEC campaigns by impersonating internal users and embedding obfuscated lures such as QR codes and empty‑body messages. Talos recommends a phased approach — inventorying dependencies, migrating devices to authenticated SMTP or partner connectors, and validating mailflows before enabling RejectDirectSend — to reduce risk without disrupting critical workflows.

Snappybee, Citrix Flaw Used to Breach European Telecom

🔒 A European telecommunications organization was targeted in the first week of July 2025, according to Darktrace, with a threat actor linked to the China-associated group Salt Typhoon gaining initial access via a vulnerable Citrix NetScaler Gateway. The intruders pivoted to Citrix VDA hosts in an MCS subnet and used SoftEther VPN to mask their origin. They deployed Snappybee (aka Deed RAT) via DLL side-loading alongside legitimate antivirus executables; the backdoor called home to aar.gandhibludtric[.]com. Darktrace says the activity was detected and remediated before significant escalation.

China Alleges NSA Cyberattack on National Time Service

🔍 China’s security authorities publicly accused the US National Security Agency of a covert operation against the National Time Service Center, alleging an SMS-service vulnerability was exploited beginning March 25, 2022 to compromise staff phones and steal data. Experts told CSO the claim is technically plausible but there is no public forensic evidence to confirm it conclusively. The alleged intrusion could affect Beijing Time, potentially disrupting communications, finance, power, transportation and space operations. Security specialists recommend hardening time infrastructure, avoiding SMS-based privileged logins, validating clocks against multiple trusted references, deploying cryptographic attestation for time signals, and following guidance from CISA.

Meta Adds Scam Warnings to WhatsApp and Messenger Apps

🔒 Meta is rolling out new anti-scam features for WhatsApp and Messenger. On WhatsApp, users will receive warnings when attempting to share their screen with unknown contacts during video calls to help prevent accidental exposure of bank details or verification codes. On Messenger, an opt-in Scam detection setting flags potentially suspicious messages from unknown senders; detection runs on-device to preserve end-to-end encryption unless users choose to submit recent messages for AI review, which removes E2EE. Meta also said it has taken action against thousands of impersonating pages and disrupted millions of accounts tied to organized scam centers.

Microsoft fixes USB input bug that broke WinRE access

🔧 Microsoft released an out-of-band cumulative update, KB5070773, to restore USB mouse and keyboard functionality in the Windows Recovery Environment (WinRE) after October 2025 security updates disabled USB input in recovery on affected client and server builds. The patch began rolling out on October 20, 2025 and Microsoft recommends installing the latest updates. If a device cannot boot to install the patch, workarounds include using a touchscreen’s touch keyboard, connecting PS/2 peripherals, or booting from a previously created USB recovery drive.

Amazon Connect Adds Automated Triggered Evaluations

🔔 Amazon Connect can now automatically initiate follow-up evaluations when specific conditions are detected during initial Contact Lens reviews. For example, if the first evaluation surfaces customer interest in a product, Connect can trigger a targeted follow-up focused on the agent's sales performance. Managers gain consistent standards across cohorts and capture deeper insights into sales opportunities, escalations, and other critical interaction moments. The capability is available in all regions where Amazon Connect is offered.

Google: Three New COLDRIVER Malware Families Identified

🔍 Google Threat Intelligence Group (GTIG) reports three new malware families — NOROBOT, YESROBOT, and MAYBEROBOT — linked to the Russia-attributed COLDRIVER group following public disclosure of LOSTKEYS. The attacks use ClickFix-style HTML lures and fake CAPTCHA prompts to trick users into running malicious PowerShell via the Windows Run dialog. NOROBOT functions as a loader invoked by rundll32.exe, while YESROBOT acted as a brief HTTPS-based Python backdoor and MAYBEROBOT is a more extensible PowerShell implant targeting high-value victims.

Developers of Lumma Stealer Doxxed in Rival Campaign

🔍Lumma Stealer operations have been disrupted after an underground doxxing campaign exposed personal and operational details of individuals allegedly tied to the malware’s development and administration. Trend Micro links the exposure to rival cybercriminal actors and reports that leaked data—shared on a site called Lumma Rats—included passports, bank details and contact information. The disclosures coincided with reduced C2 activity and the reported compromise of Telegram accounts, prompting many users to seek alternatives such as Vidar and StealC.

Google Migrates ISAs with AI and Automation at Scale

🔧 Google details how its custom Axion Arm CPUs and a mix of automation and AI enabled large-scale migration from x86 to multi-architecture production across services such as YouTube, Gmail, and BigQuery. The team analyzed 38,156 commits (about 700K changed lines) and reports migrating more than 30,000 applications to Arm while keeping both Arm and x86 in production. Existing automation like Rosie, sanitizers, fuzzers, and the CHAMP rollout framework handled much of the work, while an LLM-driven agent called CogniPort fixed build and test failures, showing a 30% success rate on a 245-commit benchmark. Google plans to default new apps to multiarch and continue refining AI tools to address the remaining long tail.

Amazon Corretto October 2025 LTS Security Updates Released

🔔 Amazon released quarterly security and critical updates for Amazon Corretto LTS builds on October 21, 2025, providing new binaries for Corretto 25.0.1, 21.0.9, 17.0.17, 11.0.29 and 8u472. The distributions for Generic Linux, Alpine and macOS now include Async-Profiler, a low‑overhead sampling profiler that captures CPU, heap and native allocations, contention and hardware/software counters. Downloads are available from the Corretto home page or via Apt, Yum and Apk repositories, and contributors can provide feedback on the Corretto GitHub.

Updates enforce SID checks, causing Windows login failures

🔒 Microsoft confirmed that Windows updates released on and after August 29, 2025 enforce additional SID checks that can break Kerberos and NTLM authentication on devices with duplicate Security Identifiers (SIDs). Affected systems — including Windows 11 24H2, Windows 11 25H2, and Windows Server 2025 — may experience failed Remote Desktop sessions, SEC_E_NO_CREDENTIALS event errors, and "access denied" messages. The fault commonly arises when images are duplicated without using Sysprep. Microsoft recommends rebuilding impacted machines with supported imaging procedures or obtaining a temporary Group Policy from Support as an interim measure.

Sophisticated Investment Scam Impersonates Singapore Official

🔍 Cybersecurity researchers have uncovered a large-scale investment scam that impersonated Singapore’s top officials, including Prime Minister Lawrence Wong and Minister K Shanmugam, to promote a fraudulent forex platform. The campaign used verified Google Ads, hundreds of fake news domains and deepfake videos, funneling victims through multiple redirects to a Mauritius-registered trading site. Group-IB reported advanced evasion techniques and localized targeting to show scam pages only to Singaporean users, pressuring many to invest and then blocking withdrawals.

Pro-Russia Information Operations After Drone Incursion

🔎 Google Threat Intelligence Group (GTIG) observed coordinated pro-Russia information operations responding to reported Russian drone incursions into Polish airspace on Sept. 9–10, 2025. Actors amplified narratives denying Russian culpability, blaming NATO or Poland, and seeking to erode domestic and international support for Ukraine. GTIG documented activity across multiple networks and languages and noted these operations leveraged both long-standing and recently developed influence infrastructure.

Ransomware Payouts Rise to $3.6M as Tactics Evolve

🔒 The average ransomware payment climbed to $3.6m in 2025, up from $2.5m in 2024, as attackers shift to fewer but more lucrative, targeted campaigns. ExtraHop's Global Threat Landscape Report found 70% of affected organisations paid ransoms, with healthcare and government incidents averaging nearly $7.5m each. The study highlights expanding risks from public cloud, third‑party integrations and generative AI, and urges organisations to map their attack surface, monitor internal traffic for lateral movement and prepare for AI‑enabled tactics.

AI-Enabled Ransomware: CISOs’ Top Security Concern

🛡️ CrowdStrike’s 2025 ransomware survey finds that AI is compressing attacker timelines and enhancing phishing, malware creation, and social engineering, forcing defenders to react in minutes rather than hours. 78% of respondents reported a ransomware incident in the past year, yet fewer than 25% recovered within 24 hours and paying victims often faced repeat compromise and data theft. CISOs rank AI-enabled ransomware as their top AI-related security concern, and many organizations are accelerating adoption of AI detection, automated response, and improved training.

Amazon Bedrock Data Automation Adds Video, Faster Images

🎞️ Amazon Bedrock Data Automation now supports AVI, MKV, and WEBM video formats and the AV1 and MPEG-4 Visual (Part 2) codecs, expanding coverage for archival, multi-track, and web-based videos. The service also delivers up to 50% faster image processing to accelerate extraction of visual insights. BDA is available in eight AWS Regions, enabling organizations to process native formats and streamline GenAI workflows.

Ransomware Reality: High Confidence, Low Preparedness

⚠️ The CrowdStrike State of Ransomware Survey reveals a sizable gap between organizational confidence and actual ransomware readiness. Half of 1,100 security leaders say they are "very well prepared," yet 78% were attacked in the past year and fewer than 25% recovered within 24 hours. The report warns that AI-accelerated attacks deepen this gap and recommends AI-native detection and response such as Falcon to regain the advantage.

Amazon SES adds IP observability for DIP-M pools capability

📬 Amazon Simple Email Service (SES) now exposes the exact IP addresses used by Dedicated IP Addresses - Managed (DIP-M) pools. Customers can view these IPs via the console, CLI, or SES API and access Microsoft SNDS metrics for each address. SES also creates CloudWatch metrics for SNDS data to aid reputation monitoring. This gives customers greater transparency into sending activity and helps diagnose deliverability and reputation issues with mailbox providers.

Fortinet Publishes First EPD for FortiGate-40F NGFW

🌱 Fortinet has published the industry’s first Environmental Product Declaration (EPD) for the FortiGate-40F Next-Generation Firewall, verified under the new PCR 2024:06. The EPD is based on an independent Life Cycle Assessment and discloses lifecycle impacts—carbon, energy, water, materials, and waste—providing procurement teams with standardized, third-party-validated data. Fortinet views this as an initial step and plans to extend EPD coverage across additional models to support compliance, decarbonization, and sustainable procurement.

Google abandons Privacy Sandbox, ends most cookie efforts

🍪 Google has announced it is discontinuing 11 Privacy Sandbox technologies — effectively ending most of the company’s cookie‑replacement efforts after evaluating low adoption and ecosystem feedback. The decision follows regulatory scrutiny from the UK’s Competition and Market Authority and several U.S. antitrust actions, and came after prior concessions from Google. The company says it will continue to work on privacy improvements for Chrome, Android and the web but will move away from the Privacy Sandbox branding.

Securing AI in Defense: Trust, Identity, and Controls

🔐 AI promises stronger cyber defense but expands the attack surface if not governed properly. Organizations must secure models, data pipelines, and agentic systems with the same rigor applied to critical infrastructure. Identity is central: treat every model or autonomous agent as a first‑class identity with scoped credentials, strong authentication, and end‑to‑end audit logging. Adopt layered controls for access, data, deployment, inference, monitoring, and model integrity to mitigate threats such as prompt injection, model poisoning, and credential leakage.

VirusTotal Success: SEQRITE APT Hunting Case Studies

🔎 SEQRITE's APT-Team describes how they used VirusTotal to pivot from isolated clues to comprehensive campaign mapping, tracking UNG0002, Silent Lynx, and DRAGONCLONE between May 2024 and May 2025. Their work combined malware configuration extraction, LNK metadata, code-sign certificate pivots, YARA and Sigma rules, and Livehunt queries to surface related samples and previously unreported implants. The post highlights practical hunting queries and pivots — public key and LNK-ID searches, submitter geofilters, and malware_config values — that enabled attribution and expanded detection across multiple Asian geographies.

The AI Fix #73: Gemini gambling, poisoning LLMs and fallout

🧠 In episode 73 of The AI Fix, hosts Graham Cluley and Mark Stockley explore a sweep of recent AI developments, from the rise of AI-generated content to high-profile figures relying on chatbots. They discuss research suggesting Google Gemini exhibits behaviours resembling pathological gambling and report on a Gemma-style model uncovering a potential cancer therapy pathway. The show also highlights legal and security concerns— including a lawyer criticised for repeated AI use, generals consulting chatbots, and techniques for poisoning LLMs with only a few malicious samples.

CISOs' 2025 Priorities: Data, AI, and Simplification

🔒 CSO's 2025 Security Priorities Study finds security leaders are juggling expanding responsibilities while facing greater complexity in selecting the right tools. Seventy-six percent say solution selection is more complex and 57% had trouble finding incident root causes in the past year. Top focuses are protecting sensitive data, securing cloud systems, and simplifying IT infrastructure, with 73% now more likely to consider AI-enabled security. Many plan to rely on managed service providers and maintain level budgets while driving strategic AI and governance initiatives.

SmarterX Builds Custom LLMs with Google Cloud Tools

🔍 SmarterX uses Google Cloud to build custom LLMs that help retailers, manufacturers, and logistics companies manage regulatory compliance across product lifecycles. Using BigQuery, Cloud Storage, Gemini, and Vertex AI, the company ingests, normalizes, and indexes unstructured regulatory and product data, applies RAG and grounding, and trains customer-specific models. The integrated platform empowers subject matter experts to evaluate, correct, and deploy model updates without heavy engineering overhead.

Deep Dive: BPF LPM Trie Performance and Optimization

🔍 Cloudflare investigated a production soft lockup traced to the Linux BPF LPM trie, a core data structure for IP and IP+Port longest-prefix matching. Benchmarks on 96-core AMD EPYC hardware showed lookups remain relatively fast at modest sizes, but updates, deletes and especially freeing maps degrade severely at scale, causing multi-second CPU stalls and customer packet loss. The post refreshes trie basics, presents measured results (lookups, updates, deletes, free costs), and diagnoses kernel implementation limits — notably binary child pointers, absent level compression, and allocator-induced cache and dTLB pressure — then outlines plans to upstream benchmarks and refactor toward a level-compressed multibit trie to reduce traversal height, cache/TLB misses, and freeing overhead.

Microsoft fixes bug blocking classic Outlook startup

🛠️ Microsoft has implemented a fix for a major issue that prevented some Microsoft 365 customers from launching the classic Outlook client on Windows. Affected users reported errors indicating the app could not be started, the Outlook window would not open, or Exchange sign-in failed. Microsoft marked the incident as fixed and said the Outlook team is monitoring the rollout, while recommending Outlook Web Access or the new Outlook for Windows as temporary workarounds.

Digital Sovereignty Sessions at AWS re:Invent 2025 Guide

📘 The AWS re:Invent 2025 attendee guide highlights the conference's digital sovereignty program, detailing sessions, workshops, and code talks focused on data residency, hybrid and edge deployments, and sovereign infrastructure. Key topics include the AWS European Sovereign Cloud, AWS Outposts, Local Zones, and security features such as the Nitro System. Practical workshops and chalk talks demonstrate RAG, agentic AI, and low-latency SLM deployments with operational controls and compliance patterns. Reserve seating via the attendee portal or access sessions with the free virtual pass.

Researchers Exploit 34 Zero-Days at Pwn2Own Ireland

🔒On the first day of Pwn2Own Ireland 2025, security researchers exploited 34 unique zero-day vulnerabilities and collected $522,500 in cash awards. Team DDOS (Bongeun Koo and Evangelos Daravigkas) chained eight flaws to compromise a QNAP Qhora-322 router via its WAN interface and access a QNAP TS-453E, earning $100,000 and moving into second place on the Master of Pwn leaderboard. The Summoning Team led day one with $102,500 and 11.5 points after multiple successful root exploits. The Zero Day Initiative (ZDI) organized the event and coordinates 90-day responsible disclosure with affected vendors.

DeepSeek Privacy and Security: What Users Should Know

🔒 DeepSeek collects extensive interaction data — chats, images and videos — plus account details, IP address and device/browser information, and retains it for an unspecified period under a vague “retain as long as needed” policy. The service operates under Chinese jurisdiction, so stored chats may be accessible to local authorities and have been observed on China Mobile servers. Users can disable model training in web and mobile Data settings, export or delete chats (export is web-only), or run the open-source model locally to avoid server-side retention, but local deployment and deletion have trade-offs and require device protections.

Maximizing Gateway Security Beyond Basic Configuration

🛡️ This article by Andrius Buinovskis of NordLayer explains why default gateway setups often leave gaps in security, performance, and compliance. It recommends four core actions: network segmentation, multiple distributed gateways to avoid single points of failure, optimization for geographically dispersed workforces, and layered cloud firewall controls to restrict ports and protocols. The guidance aligns with Zero Trust principles and highlights regional privacy rules such as GDPR and CCPA.

Scouting America Introduces Cybersecurity Merit Badge

🛡️ Scouting America (formerly Boy Scouts) has introduced a new cybersecurity merit badge that highlights digital safety, basic cyber hygiene, and introductory technical skills for youth. The announcement includes a well-designed badge image that has been picked up by mainstream coverage, drawing attention to how organizations are teaching online risk awareness. The author notes the image looks good and expresses a personal wish to earn the badge.