All news in category “Vendor and Hyperscaler Watch”

🔔 Amazon ECS now supports non-critical Managed Daemons for ECS Managed Instances, allowing operators to mark daemons so their failure won’t interrupt mission-critical application tasks. This ensures container instances remain active, existing application tasks continue running, and ECS will still place new tasks even when a daemon fails. Events are emitted to EventBridge and service action logs record daemon health; disablement is configured by setting the critical parameter to false via Console, CLI, CloudFormation, or SDKs.

🛡️ OpenAI announced Daybreak for Frontline Defenders, a $1 billion global initiative to expand subsidized access to its Daybreak cyber models, training, and technical support for critical infrastructure defenders. The program combines frontier defensive models with the Codex harness and Codex Security to identify and remediate vulnerabilities, and will integrate with enterprise tools via the Daybreak Defense Network. A U.S.-focused pilot with MS-ISAC will train public-sector and water-system defenders, while OpenAI convenes utilities and offers targeted support to affected states and operators.

🔍 This post examines a multi‑stage attack that begins with a web application SSRF vulnerability on an EC2 instance, leads to IMDSv1 credential harvesting for an attached webdev role, and culminates in unauthorized access to Amazon Bedrock across Regions. It walks through four CloudTrail events—failed CreateUser, console sign‑in without MFA, ListFoundationModels in another Region, and a Converse call invoking Amazon Nova Pro—and shows which log fields reveal attribution, intent, and cross‑Region pivots. The article also provides containment, remediation, and hardening recommendations.

🔍 This guide from the AWS Security Incident Response Team explains how to analyze AWS CloudTrail events to investigate cross-account unauthorized access, cryptocurrency mining deployments, and AI service abuse. It walks through real-world scenarios showing which CloudTrail fields matter, how to interpret session metadata, and investigative techniques for reconstructing activity timelines and assessing blast radius. The guide emphasizes practical steps to prioritize containment, identify misconfigurations such as overbroad cross-account roles or missing MFA, and extract evidentiary details from CloudTrail and related logs.

🔍 Cloudflare is offering early access to Vulnerability Discovery and Remediation, a managed service that scans authorized customer codebases using OpenAI Daybreak models (including GPT-5.6 Cyber) to find and prioritize vulnerabilities. The service correlates source findings with production traffic, WAF signals, and security events to produce prioritized fixes and scoped WAF mitigations. Customers review proposed patches and rules before any change is made.

🚀 Starting today, Amazon EC2 M9g and M9gd instances powered by AWS Graviton5 processors are available in Europe (Ireland) and Asia Pacific (Singapore, Sydney, Tokyo). These instances deliver up to 25% better performance versus Graviton4-based M8g/M8gd and up to 30% faster database performance. Built on the sixth-generation AWS Nitro System, they introduce the Nitro Isolation Engine for mathematically assured workload isolation. M9gd adds local NVMe SSD storage for low-latency needs and both instance families are available via standard purchasing options.

🚀 Starting today, Amazon EC2 P6-B200 instances accelerated by NVIDIA Blackwell GPUs are available in the AWS Asia Pacific (Hyderabad) Region. These instances deliver up to 2x performance versus P5en for AI training and inference and include 8 Blackwell GPUs, 1440 GB of high-bandwidth GPU memory, and 60% greater GPU memory bandwidth. They use 5th Gen Intel Xeon processors (Emerald Rapids), offer up to 3.2 Tbps EFAv4 networking, and run on the AWS Nitro System for scalable UltraClusters.

🚀 Starting today, Amazon EC2 P6-B300 instances are available in the AWS Asia Pacific (Jakarta) Region. These instances provide 8x NVIDIA Blackwell Ultra GPUs with 2.1 TB high-bandwidth GPU memory, 6.4 Tbps EFA networking, 300 Gbps ENA throughput, and 4 TB system memory. P6-B300 delivers higher networking and memory than P6-B200, enabling faster training and greater token throughput for large foundation models.

🔒 Microsoft frames Azure as a unified platform enabling enterprise AI in production, emphasizing integration across models, infrastructure, data, applications, and developer tools. The post highlights Azure’s multi-model support, Microsoft Foundry for model choice and operations, and data governance via Microsoft Fabric and Purview. It cites Gartner and Forrester Leader recognitions and customer examples showing modernization and responsible AI deployment.

🚀 GPT-6 Astra is now generally available to Microsoft Foundry customers, positioned to move enterprise AI from chat experiences to agentic work that reasons, plans, and executes across applications and systems. Foundry integrates identity, networking, governance, and monitoring to help organizations deploy Astra with scoped credentials, human checkpoints, and activity records. Deployment options include Standard and Provisioned Throughput across Global and US Data Zone geographies, with pricing tiers for short and long contexts.

🚀 Customers can now subscribe and manage CloudFront flat-rate pricing plans programmatically via the AWS CLI, SDKs, CloudFormation, CDK, or the PricingPlanManager API. Flat-rate plans cover global content delivery, WAF, DDoS, DNS, logging, and edge compute under one monthly fee without usage overages. Paid plans offer a two-phase activation (create then approve) for controlled billing, while free plans activate immediately. There are no extra fees for using the API.

🚀 Amazon Quick introduces Quick Max, a new plan aimed at power users, offering 5x the usage and 5x the storage of the Plus tier. The plan supports large, concurrent workloads throughout the month and provides greater value at higher usage levels. Quick Max is available with monthly and annual billing; existing Plus users can upgrade via the product navigation. New users can sign up free, and plan comparisons are available on the pricing page.

🔧 AWS Gateway Load Balancer (GWLB) now supports sending TCP Reset packets to accelerate failure recovery for client and server connections. Previously, GWLB exhibited fail-open behavior where traffic continued to be forwarded to unhealthy targets, causing prolonged interruptions due to TCP retries and back-off. TCP Reset can be enabled per target group via the Console, CLI, or API and responds to three independent triggers: target unhealthy, target deregistration after connection draining, and TCP idle timeout expiry. This capability is available in all GWLB regions at no additional cost and is off by default for backward compatibility.

🖥️ Amazon WorkSpaces Applications now supports Graphics G7 instances powered by NVIDIA RTX PRO 4500 Blackwell Server Edition GPUs and Intel Xeon Scalable (6th Gen) processors. G7 delivers up to 2.1× better performance for graphics workloads versus G6 and offers 32 GB of GDDR7 per GPU with 2.67× faster memory bandwidth. Six sizes are available (1–8 GPUs, 8–192 vCPUs, 32–768 GB RAM) and G7 is initially available in three US regions.

🔷 Amazon Redshift now supports single-node clusters for rg.large instances on P204 or later patches, enabling cost-effective testing and proof-of-concept deployments without high-availability requirements. RG instances, powered by AWS Graviton processors, offer up to 2.4x faster performance than prior RA3 generations and a ~30% lower price per vCPU. The RG family includes Redshift’s custom vectorized engine for SQL analytics over Apache Iceberg and Parquet data, and these instances are available across numerous global AWS Regions.

🔒 Amazon Aurora MySQL-Compatible Edition 8.4 now supports MySQL 8.4.8, introducing security enhancements and bug fixes plus features such as post-quantum TLS (PQ-TLS) key exchange, transaction timeout, multi-source replication, and delayed replication. These additions strengthen in-transit encryption options and improve operational resilience by preventing long-running transaction impacts and enabling consolidated or lagged replicas for recovery and reporting. Upgrades are available via automatic minor version upgrades during scheduled maintenance and supported across all AWS Regions where Aurora MySQL is offered.

📢 Amazon Aurora MySQL now supports multi-source replication and delayed replication, enabling consolidation of multiple MySQL sources into a single Aurora cluster and intentional replica lag for protection against human error and logical corruption. Multi-source replication simplifies merging shards or aggregating regional and departmental databases into a central cluster for reporting and backups. Delayed replication allows a binlog replica to intentionally lag, letting you stop replication and promote the replica to recover quickly from harmful changes without a full restore. These features are available on Aurora MySQL 8.4.8+ in all Regions where Aurora MySQL is offered.

🐭 Microsoft confirmed that a bug introduced by the KB5120998 August 27, 2026 preview update reverts mouse personalization settings on some Windows 11 devices. The issue was found to affect only non‑English installations, where necessary code components fail to load and default cursor settings are applied. KB5120998 is an optional preview update that can install automatically if users opted to receive updates as soon as available. Microsoft has acknowledged related mouse and desktop settings issues and continues to investigate and address them.

🛠️ OpenAI confirmed a widespread ChatGPT and Codex outage that began around 10:58 AM ET on Thursday, September 3. The disruption affects Conversations, Login, ChatGPT Work, GPTs, image generation, Voice mode, file uploads, Deep Research, Agents and other components; at least 15 services are listed as impacted. OpenAI acknowledged the issue on its status page and says it is investigating while continuing to monitor elevated errors.

🔒 Fortinet’s partnership with the Forum of Incident Response and Security Teams (FIRST) advances global cyber resilience by combining FortiGuard Labs’ threat intelligence with FIRST’s incident response network. CORE, a 2025 initiative co-founded by Fortinet, funds capacity building, regional training, and tabletop exercises to close skills gaps and improve cross-border coordination. This collaboration also supports standards like EPSS and CVSS to turn shared practices into operational defenses.