India Microsoft designated its Hyderabad-based India South Central cloud region as a strategic hub for Asia and the Global South and said all its Indian cloud regions now have AI-capable hardware. The Hyderabad region has three availability zones and forms part of Microsoft's previously announced US$20.5 billion India cloud and AI investment commitment. Microsoft named Adani Group, Bajaj Finserv, HDFC Bank and PB Pay among customers signing up for the region. The region itself became generally available on 6 August 2026; the in-window development was the 21 September strategic-hub designation and broader AI-enablement announcement.
India Tech Mahindra and CoRover.ai announced a partnership combining CoRover's BharatGPT, Tech Mahindra's Project Indus and TechM Orion. The companies intend to target sovereign AI, agentic AI, enterprise applications and digital public infrastructure, including government and citizen services. Deployment options span edge, on-premises, cloud and hybrid environments, with an emphasis on Indian languages and locally relevant context. No quantified customer deployment or business outcome was announced.
India Brahma AI announced a US$150 million preferred-share financing led by Multiples Alternate Asset Management to expand R&D and international commercialization of its enterprise audiovisual AI platform. The announced round includes US$100 million from Multiples. Filing-based reporting adds an important qualification: the broader $150 million represented the targeted round size, while not all additional investor interest had necessarily become completed subscriptions at announcement.
Saudi Arabia Saudi Arabia's Ministry of Commerce reported 24,552 commercial registrations for AI activities, representing 31% growth versus 2025 and the highest growth rate among the digital/Vision 2030 sectors highlighted by the ministry. Other reported totals included 11,275 cybersecurity registrations (+22%), 13,853 VR/AR registrations (+21%) and 52,241 e-commerce registrations (+9%). These figures measure commercial registrations, not funded startups, active AI products or enterprise deployments.
Saudi Arabia Saudi enterprise-AI company MOZN announced partnerships with NourNet, MDS and Ejada at Money20/20 Middle East. NourNet contributes cloud and datacenter infrastructure, MDS provides regional systems integration and delivery, and Ejada adds integration capabilities particularly relevant to banking and financial services. The agreements focus on implementation, co-selling and secure enterprise deployment rather than announcing a specific new production customer or quantified outcome.
United States Texas Governor Greg Abbott directed the Texas Commission on Environmental Quality to halt permits sought by data centres until ERCOT's audit is completed. The state is examining electricity demand, water consumption, infrastructure costs, project ownership and grid reliability. ERCOT had approximately 474 GW of connection requests under consideration in August, with roughly 90% associated with data centres, according to the governor's earlier audit directive.
United States Palo Alto Networks introduced Unit 42 Continuous Frontier AI Defense, an agentic offensive-security service intended to continuously identify, validate and remediate enterprise exposures. The company says the service uses gated-capability models including Anthropic Claude Mythos and OpenAI GPT models. Palo Alto reported that AI-enabled attackers have compressed some breach cycles by almost 97%, but that statistic is a company-reported security observation rather than an independently established market measure.
United States Okta announced the Blueprint Alliance, bringing together companies including AWS, CrowdStrike, Databricks, Docker, Google Cloud, Okta, Proofpoint, Salesforce, ServiceNow, Wiz and Zscaler. The initiative targets a shared architecture for securing enterprise AI agents across identity, authorization, credentials, applications and infrastructure. The development is significant because agent security is moving beyond model-level safeguards toward interoperable enterprise identity and access controls.
United States UiPath announced platform enhancements spanning agent governance, enterprise data, connectivity and deployment, alongside UiPath Cartographer. Cartographer creates a living, governed “Map of Work” designed to represent how business processes operate and provide contextual grounding for agents and automation. UiPath also announced AI-driven autonomous-testing capabilities, extending its platform beyond task automation toward governed agentic operations.
United States VAST Data introduced DataEnclave, a confidential-AI runtime built on NVIDIA Confidential Computing. The architecture is designed to allow proprietary and open models to operate against sensitive enterprise data inside customer datacentres or trusted cloud hardware while protecting both the organization's data and model intellectual property. The announced ecosystem includes Cohere, CrowdStrike, Deepgram, Factory, Fundamental, NVIDIA and TwelveLabs.
United States Oracle Health announced five planned AI capabilities covering prior authorization, clinical-documentation integrity, charge capture, professional-fee coding and appeal management. Oracle intends to embed these capabilities into its existing revenue-cycle products and release them over coming months. The significance is the placement of AI directly inside healthcare transaction workflows where documentation or coding errors can subsequently produce denials and payment delays; no quantified financial outcome has yet been established.
United States INRY and ServiceNow reported that INRY put ServiceNow EmployeeWorks into production in six weeks, replacing a legacy employee portal with a conversational interface. INRY was both an early-access customer and a ServiceNow implementation partner. The six-week figure therefore represents a jointly reported customer implementation result rather than an independent benchmark of typical deployment speed.
European Union The European Commission proposed a common rating scheme intended to make EU data-centre energy and water performance transparent and comparable. It also opened consultation on possible minimum performance standards, with a legislative proposal planned for Q2 2027. The Commission says Europe aims to triple data-centre capacity over roughly five to seven years, making resource efficiency increasingly relevant to AI-infrastructure policy and procurement.
France BNP Paribas and Google Cloud signed a five-year agreement expanding the bank's access to Google Cloud AI infrastructure, Gemini models and Gemini Enterprise. Initial planned applications include integrating Gemini into Corporate & Institutional Banking's LLM@CIB assistant and deploying agentic-AI use cases. BNP Paribas says the work will operate within its existing multi-cloud architecture and security and data-governance controls.
Germany SAP unveiled capabilities across Signavio, LeanIX, WalkMe and Cloud ALM aimed at governing AI agents throughout their lifecycle. Key additions include Company Memory for organizational rules and process knowledge, an AI Governance Assistant incorporating EU AI Act and NIST compliance intelligence, a UI-native agent capable of interacting with legacy interfaces, and Cloud ALM monitoring of agent actions in production. The architecture directly connects enterprise context, agent execution, compliance and operational monitoring.
China Alibaba unveiled its Zhenwu V900 AI chip, which it says delivers 3× the performance of the previous M890 and can participate in clusters scaling to 500,000 cards. CEO Eddie Wu said Alibaba is training Qwen 4 and plans future models in the 5–10 trillion parameter range. The company's infrastructure roadmap also points toward Alibaba Cloud datacentre capacity exceeding 20 GW by 2032. Chip-performance and future-scale numbers remain company claims.
Singapore Singapore-based Sea announced that it will become the first enterprise in Southeast Asia to adopt NVIDIA's Vera Rubin platform. Sea says the infrastructure will support development of AI models and intelligent agents across its businesses at greater scale. The announcement did not disclose the investment value, number of systems/GPUs, deployed compute capacity or a quantified business outcome.
