Human
Anthropic launched Project Glasswing, a defensive cybersecurity alliance featuring Amazon, Apple, and Google, powered by the unreleased 10-trillion parameter Claude Mythos model. The initiative aims to use AI to identify and patch vulnerabilities across critical hardware and software infrastructure at an unprecedented scale, moving beyond human-bound security capacity to address sophisticated emerging digital threats across the global software supply chain.
Autonomous driving leader Pony.ai officially launched Europe’s first commercial robotaxi service in Zagreb, Croatia, in partnership with Verne and Uber. Covering 90 square kilometers including the city center and airport, the fleet operates daily via a dedicated app. This deployment validates the commercial readiness of Gen-7 autonomous tech and marks a significant international expansion of robotaxi infrastructure into the European market.
The PyTorch Foundation officially adopted Safetensors, a secure tensor serialization format developed by Hugging Face, as a primary foundation-hosted project. By preventing arbitrary code execution and optimizing multi-GPU performance, this move establishes a new security standard for AI model distribution. It addresses the critical need for trust and safety within the open-source machine learning ecosystem as models grow in complexity.
Morningstar DBRS released an intelligence report detailing massive 2026 regulatory shifts aimed at managing the exponential surge in electricity demand from AI data centers. With power requirements projected to reach 106 gigawatts by 2035, new federal and state reforms are being introduced to balance grid reliability with ratepayer protection. This forces hyperscalers to take greater financial accountability for grid stability.
Timekettle debuted its W4 AI Interpreter Earbuds at GITEX Asia, introducing a hardware-native "Babel OS 2.0" designed for real-world business communication. Utilizing bone-conduction voice pickup and a dynamic engine selector, the device solves the long-standing challenge of high-noise environment speech capture. This represents a leap in specialized AI hardware for seamless, cross-lingual professional collaboration in diverse global trade settings.
