Microsoft: ‘Use Copilot at your own risk’

⚠️ Microsoft: ‘Use Copilot at your own risk’ LINK

🚨 Iran threatens OpenAI’s $30B ‘Stargate’ data center LINK

📚 Nature: AI-spawned fake citations flood 2025 literature LINK

🐉 BBC: China’s OpenClaw frenzy and its AI ambitions LINK

📈 CNBC: China’s chip firms hit record revenue on AI boom LINK

➕ 11 other news & articles you might like

🧰 5 trending tools

📚 3 trending papers & reports

TechSpot highlights Microsoft’s disclaimer that users assume the risk of Copilot’s mistakes and should verify outputs.

The cautionary language contrasts with Copilot’s deep push across Windows, Office, and enterprise workflows.

It underscores the growing tension between rapid AI rollout and the need for reliability, compliance, and accountability.

State media shared satellite imagery of the facility alongside the threat, spotlighting AI infrastructure as a geopolitical target.

A direct strike on a hyperscale AI hub would carry global ramifications for compute capacity and regional stability.

The episode raises urgent questions around physical security and redundancy for critical AI data centers.

A Nature analysis suggests tens of thousands of 2025 publications may include invalid, AI-generated references.

Editors and institutions are urged to require disclosure of AI use and deploy automated checks to detect fabricated citations.

Left unchecked, citation pollution undermines peer review, meta-analyses, and cumulative scientific progress.

BBC explores the rapid uptake of OpenClaw, an open-source agent framework drawing intense developer interest in China.

The momentum reflects demand for domestic, open alternatives amid tightening controls and ecosystem fragmentation.

It showcases how open tooling can catalyze national-scale coordination in AI development.

CNBC reports Chinese semiconductor firms hit all-time highs, fueled by AI workloads and import substitution.

U.S. export restrictions appear to be accelerating domestic capacity building and supply-chain realignment.

The surge intensifies global competition for memory, accelerators, and advanced packaging.

In Japan, robots are filling the jobs no one wants LINK

Artemis II laser link beams 4K video from the Moon LINK

CBP facility codes reportedly leaked via online flashcards LINK

Target warns shoppers may bear costs of AI agent mistakes LINK

Iran’s internet shutdown is now the longest in world history LINK

Anthropic paper stirs debate on anthropomorphizing AI LINK

Researchers train living rat neurons for real-time AI computations LINK

Drones are beating missiles; AI is doing the same to incumbents LINK

Apple said to be at a ‘fork in the road’ as it races to rebuild Siri LINK

AI-powered cameras on Wichita school buses spur privacy concerns LINK

Samsung Messages app to be discontinued in July 2026 LINK

OpenClaw + Gemma 4 (Guide): Step-by-step setup for a fully local, self-hosted AI agent stack via Ollama—no cloud or API fees. LINK

design-ai: A starter design system (tokens, styles) that AI coding agents can follow to build on-brand UIs. LINK

Linux 7.0-rc7 AI Bug Report Docs: Kernel documentation updates aimed at improving AI-generated security bug reports. LINK

Strait Watch Dashboard: Live AIS tanker traffic and ‘Days Until Dark’ oil cover tracker focused on the Strait of Hormuz. LINK

Claude Usage Limits (Update): Overview of Anthropic’s new usage caps and practical tips to work within them. LINK

Nature analysis finds tens of thousands of 2025 papers may contain AI-fabricated citations, urging stronger detection and disclosure. LINK

Researchers demonstrate living rat neurons performing real-time AI computations, hinting at bio-hybrid brain–machine interfaces. LINK

An AI-powered robot improves tomato harvesting efficiency through learned perception and control in real-world tests. LINK