On August 27, 2026, more than 100 companies and other organizations warned that AI-enabled attacks could scale within months, but their cyber-defense letter set no binding duties, budgets or deadlines.
MTG-I2 launched on August 27, 2026, completing the first MTG trio in orbit. After commissioning, it will add 2.5-minute European scans for severe-weather nowcasting.
On August 26, active exploitation was documented against qualifying Zimbra servers. Systems below 10.1.20 need an exposure check, urgent patching, and compromise review.
AWS and Nvidia unveiled plans on August 26 for 2 million additional GPUs across AWS infrastructure in 2027–2028, following a 1 million-plus commitment roughly five months earlier.
CISA added an actively exploited Gitea RCE to its catalog on August 25, 2026. Systems before 1.27.1 need patching, while open registration can give outsiders the required repository access.
On August 27, Google unveiled a test that could take viewers from YouTube Demand Gen ads into brand chats; travel changes are sector-specific, while its multimodal video tool is generally available.
Nutanix made NAI 2.8 generally available on August 26 with a GA gateway for governing MCP access. Nutanix Kubernetes Platform 2.19 remains forthcoming, without a firm release date.
Rackspace Cloud launched in Dallas on August 25, 2026, as a managed multitenant VCF 9.1 service. North Virginia and Chicago remain planned, with no opening dates disclosed.
Cisco is adding Supermicro rack-scale compute and cooling to its NVIDIA-backed architecture, with ordering due to open in October 2026. A universal October shipment date has not been published.
On August 25, WhatsApp unveiled a security rollout replacing its six-digit verification PIN with a full password, while adding multiple passkeys and Android-only caller context.
On August 25, 2026, Anthropic began sharing Claude’s memory between Chat and cloud Cowork tasks. Saved topics are editable, while sensitive categories require separate consent.
On August 25, OpenAI published leading vendor-reported InferenceX results for Jalapeño, an inference accelerator that serves trained models but cannot train new ones.
On August 25, 2026, Stability AI disclosed a $76 million Series B backed by entertainment and technology investors, including companies already co-developing its creative tools.
QUASA (https://quasa.io/), the Web3 media and creator economy platform, is expanding far beyond English, Russian, Hindi, Spanish, French, and Chinese with fully localized editions in Arabic, Portuguese, Indonesian, Bengali, Japanese, German, Urdu, Turkish, Vietnamese, and Korean.
Roman is sealed in its Falcon Heavy fairing for a no-earlier-than Aug. 30 liftoff. NASA coverage starts at 6:20 a.m. EDT, while final launch checks remain ahead.
