AI News

Chinese AgiBot showcased a fleet of humanoids in Dubai: the receptionist A2, the general-purpose G1, the dancer X2, and the all-terrain D1 Ultra. In just three years of existence, it has become the only company with a full portfolio of robots for every scenario.

SAP and Anthropic integrate the Claude model into the Joule platform. Thanks to the open MCP standard, AI agents will be able to autonomously handle tasks in finance, HR, procurement, and logistics — without the need to build custom integrations for each system.

Singapore announced a major partnership with OpenAI, NVIDIA, and robotics firms at the ATxSummit 2026 summit. The Punggol Digital District is becoming the world's first public robot testing zone, and OpenAI is investing over 5 billion crowns into the country. While the EU fine-tunes the AI Act, the Asian tiger is betting on speed and practical deployment.

Browserbase has launched Browse.sh — an open catalog of over 100 browser skills for AI agents. A single CLI command installs a skill that tells the agent exactly how to navigate a specific website. The result is up to 45% cost savings.

Microsoft has released two open-source tools for AI agent security. RAMPART brings automated prompt injection attack testing directly into your CI/CD pipeline, while Clarity helps teams uncover design flaws before they write a single line of code. Both tools are free under the MIT license.

Deloitte and Google Cloud have released an analysis of agentic AI in businesses. They show how to bridge the gap between AI investments and real results. The key is changing the way we work, not just buying technology.

Google DeepMind introduced the Gemini Omni model at Google I/O 2026 — a multimodal AI that generates and edits video through natural conversation. The model understands physics, history, and science and is also available in Czechia.

Singapore has signed strategic agreements with Google and OpenAI. The ChatGPT-maker is investing roughly 5.4 billion Czech koruna into the local AI ecosystem and opening its first research lab outside the United States. Google, meanwhile, will focus on education, healthcare, and science.

Consulting giant Deloitte has published a detailed framework for understanding how agentic AI will reshape work in the financial sector. Rather than speculating on whether AI will replace portfolio managers, it outlines nine concrete models for human-machine collaboration.

Alibaba has introduced Qwen3.7-Max, its most powerful AI model for the agentic era. It autonomously programs, optimizes GPU kernels, and can work continuously for 35 hours with over a thousand tool calls. In benchmarks, it beats both Opus-4.6 and DeepSeek-V4.