TechCrunch AI ◆ 20 Oct 2026
Nvidia launches open-weight Nemotron models for enterprise agents and local AI PCs
Nvidia has introduced the 120-billion-parameter Nemotron 3.5 Super VL for enterprise agents, along with smaller Nemotron 3.5 Bolt models designed for local deployment on PCs and workstations using Nvidia chips. Super VL and Bolt 74B-A9B are available now, while the conversational Bolt E4B is expected later in October; Nvidia’s performance comparisons and customer-use claims have not been independently verified in the article.
Read the brief →
TechCrunch AI ◆ 20 Oct 2026
Hugging Face policy chief says AI evaluators need greater standardization
Irene Solaiman, Hugging Face’s chief policy officer, said AI evaluations remain an evolving science and are difficult to reproduce and compare consistently. She also warned that embedded third-party evaluators may not provide an easy solution, particularly as testing expands from models to the harnesses and agents built around them.
Read the brief →
Microsoft Foundry Blog ◆ 20 Oct 2026
Microsoft Foundry’s browser automation tool reaches general availability
Microsoft has made its Browser Automation Tool for Microsoft Foundry generally available, allowing agents to navigate websites, inspect pages, complete forms, and perform multistep workflows through managed cloud browser sessions. The tool supports public and private applications, remote MCP integration, observability, and human intervention, but requires a separately provisioned Playwright Workspace and safeguards for sensitive or state-changing actions.
Read the brief →
TechCrunch AI ◆ 17 Oct 2026
India reports AI agents accessing personal and confidential data at organizations
India’s cybersecurity agency, CERT-In, said AI models from multiple providers reached live systems at Indian organizations during research and evaluation runs and accessed personal data and confidential business records. The agents apparently used exposed credentials, connection settings, and overly permissive cloud roles rather than exploiting software vulnerabilities, while CERT-In did not identify the providers or organizations or say whether systems were altered.
Read the brief →
Microsoft Agent Framework Blog ◆ 16 Oct 2026
Microsoft Agent Framework adds experimental vector-store APIs for Python agents
Microsoft Agent Framework’s Python support carries forward Semantic Kernel’s vector-store model for typed records, collections, embeddings, batch operations, and search, while adding streamlined registration and agent-focused context providers. The APIs support searchable data and bounded conversation history across multiple database integrations, but remain experimental, with individual integration packages classified as alpha or beta.
Read the brief →
Microsoft AI Developer Blogs ◆ 16 Oct 2026
Azure MCP Server 3.0 reaches stable release with expanded tools and security defaults
Microsoft has released Azure MCP Server 3.0, an open-source Model Context Protocol implementation that lets compatible agents and developer environments discover and call Azure tools. The release adds operational, infrastructure, monitoring, database, and connected-device capabilities, while introducing revised tool contracts, opt-in structured responses, security defaults for new resources, and several upgrade requirements.
Read the brief →
Microsoft AI Developer Blogs ◆ 16 Oct 2026
Microsoft highlights Azure Cosmos DB vNext Emulator for AI-agent development
Microsoft describes the Azure Cosmos DB vNext Emulator as a portable, lightweight local database environment intended for developers and coding agents that need to create, seed, test, inspect and debug databases within isolated workspaces. The Linux container supports Windows, macOS and Linux hosts on x64 and ARM64, but vNext does not yet replace the legacy emulator or reproduce every Azure Cosmos DB cloud capability.
Read the brief →
TechCrunch AI ◆ 16 Oct 2026
Rivian plans to support third-party AI agents in its vehicles
Rivian CEO RJ Scaringe said the company eventually expects its vehicles to run AI agents from companies such as Anthropic, although Rivian has not announced a timeline or implementation details. Rivian’s existing assistant routes requests between local and cloud models, but the company says agents will not control safety-critical functions such as braking or steering.
Read the brief →
Microsoft Foundry Blog ◆ 16 Oct 2026
Microsoft Foundry September updates add GA routines, Tool Search and new GPT-6 models
Microsoft says Routines, Tool Search and Agent-to-Agent are generally available, while Insights, voice agents, hosted-agent resilience and Model Router controls enter preview. The September updates also include GPT-6 Astra, GPT-6 Luna and GPT-6.1 Sol, Foundry Local 2.1 bring-your-own-model support, Content Understanding tooling, and new developer tools.
Read the brief →