Pentagon Greenlights AI Model Training on Classified Data
The Pentagon is moving forward with a plan to allow technology companies to train their artificial intelligence models on classified data. This initiative represents a significant step in advancing the U.S.’s defense technology capabilities. The strategy aims to accelerate AI development for national security purposes while navigating the complex challenges of juggling innovation with the protection of sensitive information. This move highlights the increasing integration of AI into critical defense infrastructure and the need for robust, adaptive cybersecurity measures to safeguard against potential risks.
Sources:
Rivian and Uber Forge Partnership for 50,000 Autonomous Robotaxis
Electric vehicle manufacturer Rivian and ride-sharing company Uber have announced a strategic partnership to deploy up to 50,000 fully autonomous robotaxis. The collaboration will utilize Rivian’s upcoming R2 SUV platform, with the vehicles being exclusively available through the Uber network. Uber is set to invest up to $1.25 billion in Rivian by 2031, contingent on the achievement of key autonomous performance milestones. The initial rollout is planned to begin in 2028 with 10,000 autonomous R2s, starting in San Francisco and Miami, with an expansion to 25 cities by 2031. This partnership is a significant step for Rivian’s autonomous driving ambitions and expands Uber’s portfolio of self-driving vehicle suppliers.
Sources:
Elon Musk Unveils ‘Terafab’ to Mass-Produce AI Chips
Elon Musk has unveiled ‘Terafab,’ a new joint venture between Tesla, SpaceX, and xAI focused on the large-scale manufacturing of artificial intelligence chips. The facility, located near Austin, Texas, will aim to produce 100-200 billion AI and memory chips annually using 2nm process technology, with a total planned investment of approximately $20 billion. The chips produced at Terafab are intended to support Tesla’s AI development, xAI’s Grok infrastructure, and the deployment of Optimus robots. This move is seen as a strategic effort by Musk to create a vertically integrated AI hardware supply chain, reducing reliance on external chip manufacturers.
Sources:
Mistral AI Launches ‘Forge’ for Custom Enterprise AI Model Development
Mistral AI has launched ‘Forge,’ a new platform that enables enterprises to create and train their own generative AI models using proprietary data. This offering is designed to bridge the gap between general-purpose AI systems and the specific needs of businesses by allowing them to build models that understand their internal data, processes, and regulatory environments. The Forge platform supports various stages of the model lifecycle, including pre-training on internal datasets and reinforcement learning to align with company policies. Emphasizing data control, Mistral allows organizations to maintain ownership of both their models and the underlying data.
NVIDIA Debuts Open Platform for Building Enterprise AI Agents
NVIDIA has launched a new open platform designed to help businesses create enterprise-grade AI agents. This initiative is part of a broader industry trend towards more autonomous AI systems that can perform tasks rather than just generating content. The platform aims to enable the development of AI agents capable of managing workflows, analyzing data, and executing actions across various digital platforms. This move is expected to accelerate the adoption of AI in core business operations and reflects the increasing competition among major tech companies to provide the foundational tools for building sophisticated AI applications.
Sources:
Broadcom Updates VMware vSphere Kubernetes Service with Kubernetes 1.35 Support
At KubeCon + CloudNativeCon Europe 2026, Broadcom announced VMware vSphere Kubernetes Service (VKS) 3.6. This new version introduces support for Kubernetes 1.35, providing enterprise operations with extended support. VKS 3.6 is designed to address the challenges platform teams encounter after the initial deployment, focusing on safe upgrades, predictable operations, and support for demanding workloads like databases. The update also includes support for Red Hat Enterprise Linux (RHEL) 9 for both control plane and worker nodes, joining other supported operating systems like Photon OS 5, Ubuntu, and Windows Server 2022. Additionally, VKS 3.6 features centralized, API-driven management of node-level firewall rules and adds support for the nftables backend for kube-proxy on Linux nodes to enhance performance. Broadcom is also positioning its VMware Avi Load Balancer as a replacement for the now-deprecated open-source Ingress NGINX controller.
Sources:
Microsoft Foundry Integrates Azure DevOps for Enhanced AI Agent Access
Microsoft has announced the integration of the Azure DevOps Remote MCP Server into Microsoft Foundry, which is now in public preview. This integration enables AI agents built within Foundry to directly query and act upon Azure DevOps data, including work items, pull requests, pipelines, and repositories. The Remote MCP Server provides a managed alternative to building custom integrations, handling authentication, access control, and API complexity. Authentication is managed through Microsoft Entra, inheriting existing identity rules and security policies. This allows developers to focus on workflow logic while utilizing Foundry’s governance and security controls for deploying AI-powered applications and agents at scale.
Sources:
Tenable and OX Security Partner to Unify Cloud and Application Security
Tenable and OX Security have announced a new integration aimed at providing unified visibility across cloud and application environments. The partnership connects Tenable’s cloud security findings with OX Security’s application-level context, mapping risks directly to the relevant code and developer ownership. By synchronizing via API, Tenable Cloud Security identifies risks in cloud environments, which OX Security then enriches by pinpointing the exact application, service, and line of code responsible. The solution also includes reachability and exploitability analysis to validate if the identified risks are exposed in production, helping to reduce alert fatigue for security teams. This collaboration aims to accelerate remediation by ensuring that security issues are directed to the appropriate stakeholders without delay.
Sources:
Italian Antitrust Authority Probes Quantum Computing Market for Competition Risks
The Italian Competition Authority (AGCM) has opened a formal sector inquiry into the developing quantum computing market. Citing the technology’s transformative potential in strategic areas like cybersecurity, biotechnology, and fintech, the authority’s goal is to identify potential risks to competition before the technology achieves full commercial maturity. This proactive investigation aims to understand the market’s structure and prevent potential competitive imbalances as the industry grows.
Sources:
Xiaomi Unveils MiMo-V2-Pro, a Powerful New Open-Source AI Model
Xiaomi’s AI division, MiMo, has released a new powerful open-source AI model, MiMo-V2-Pro, which was initially tested in stealth as “Hunter Alpha” on the AI gateway platform OpenRouter. This one-trillion-parameter model with a one-million-token context is designed with a focus on agentic capabilities. The release is part of a broader trend in March 2026 that sees the gap narrowing between proprietary and open-source AI models. This is also demonstrated by the continued growth of the OpenClaw ecosystem, an open-source personal AI assistant that runs on user hardware and has surpassed 210,000 stars on GitHub. The OpenClaw platform’s skill registry now features over 5,700 community-contributed skills. Xiaomi also launched two other models: MiMo-V2-Omni for multimodal agent capabilities and MiMo-V2-TTS for text-to-speech.
Open Source Community Adapts to the Rise of AI in Development
The open-source ecosystem is rapidly evolving with the integration of Artificial Intelligence. In 2025, GitHub saw the addition of 36 million new developers, with significant growth in India, Brazil, Indonesia, Japan, and Germany, making open source more global. AI is playing a major role in this global expansion. In response to these changes, The Linux Foundation has announced the schedule for the Open Source Summit North America 2026, which will feature sessions on the rise of AI agents and strengthening software supply chain security. The summit will include discussions on building and deploying production-ready AI agents and managing the increase of AI-generated code contributions. This highlights the industry’s focus on adapting to and leveraging AI while addressing its challenges.
Sources: