AI Agents

NVIDIA has unveiled its new RTX Spark PC platform at GTC Taipei, featuring 1 petaflop of AI compute and 128GB of unified memory to run secure local AI agents.

NVIDIA has launched its new Vera CPU, delivering 1.8x faster task completion than x86 processors to power AI agents in modern data centers.

NVIDIA and Microsoft have announced the RTX Spark superchip, bringing 1 petaflop of local AI performance and advanced security to Windows PCs.

Microsoft Build 2026 begins June 2 in San Francisco, featuring a keynote by Satya Nadella on AI agents and developer tools.

Claude Code has introduced dynamic workflows in a new research preview, enabling developers to run parallel subagents for complex programming tasks.

The WorkBuddy AI desktop agent has launched globally, offering automated task execution across coding, analysis, and productivity workflows.

Google announces Gemini Spark, a 24/7 personal AI agent powered by Gemini 3.5 that performs digital tasks in the background without requiring users to keep their devices active.

CODE81 releases an e-book on autonomous and agentic applications, addressing how enterprises can adopt AI Agents responsibly within core systems as governance and scale become critical factors.

Anthropic acquires Stainless to integrate SDK and MCP server tooling, strengthening Claude’s ability to connect AI agents with external systems and APIs across multiple programming languages.

Nokia agentic AI broadband solutions embed AI agents across platforms to help telecom operators improve network management, field operations, and customer service. General availability is planned for the second half of 2026.

Group-IB deployed Prevyn AI cybersecurity as a predictive system that coordinates 11 specialized agents for threat research and reduces security operations workload through automated incident response.

HPE introduces autonomous networking powered by AI agents that detect and resolve issues without human intervention, with real-world deployments showing 75% reduction in support requests.

SAS expanded its AI agents portfolio with new industry-specific solutions designed to address barriers preventing successful artificial intelligence deployment. The Supply Chain Agent and Worker Safety offerings are now available globally.

Mistral AI Work mode launches for Le Chat, upgrading the platform to function as a multi-step execution agent powered by Mistral Medium 3.5.

The OpenAI AI phone is being developed with MediaTek, Qualcomm, and Luxshare targeting 2028 mass production. The device shifts focus from app management to AI-driven task completion.

OpenClaw vs Hermes is the biggest AI agent debate of 2026. This honest breakdown explains what each platform does, where they win, and which philosophy fits your workflow.

Google Cloud launches a next-generation data lakehouse combining Apache Iceberg storage with cross-cloud interoperability and AI-powered context for enterprise agents. The platform promises 117% ROI with payback in under six months.

CIS releases comprehensive security guidance for AI agents, adapting 18 critical security controls to address autonomous systems, tool execution, memory management, and multi-step reasoning workflows.

Google has released Gemini CLI subagents, allowing developers to delegate complex tasks to specialized expert agents that operate in isolated context windows. The feature supports parallel execution and custom agent creation.

Cloudflare launched Artifacts, a Git-compatible versioned storage system designed for AI agents. The platform enables developers to create and manage millions of repositories programmatically.

Google has added an Agent tab to Gemini Enterprise with task execution features, hinting at a desktop agent product ahead of Google I/O.

Kodamai has launched a mathematically verified AI agent platform and signed First Mills as its first enterprise customer in Saudi Arabia, deploying across four production facilities.