Sydney date / AI technology
News archive
Wednesday 2 September 2026 · 31 stories
AI technology
31
Tesla FSD Supervised: regulatory review in Europe
Defence AI monitoring: Russia
Defence AI monitoring: Ukraine
AI software development - Empirik launches with US$21 million
Energy technology - Octopus Energy acquires BD Energy
SEMICON Taiwan: semiconductor policy positioning
Anthropic hardened its evaluation and training systems after agent incidents
Defence AI monitoring: Israel
Cybersecurity: AI-enhanced attack pressure on the energy sector
ESA wildfire Earth observation: onboard AI
Tesla releases European FSD safety dashboard ahead of possible approval vote
WiFlow: Optical Flow from Wi-Fi Channel State Information
Atlas: On-Device City-Scale 3D Gaussian Splatting in VR
Language Models Can Control Their Own Attention
Efficient GUI Agents: A Systems Survey
EarlyEval: Cheaper Agent Evaluation via Early Outcome Prediction
Beyond Outcome Gaps: Process-Aware Fairness Diagnosis for LLM-based Multi-Agent Decision Systems
Stanford Online offers a course on AI agents and evaluation
LaMoC: Loss-Aware Modular Compression for LLMs
Improving Information Extraction with Learned Queries
Auditing MCQA Benchmarks through Probability Landscapes
United States presses a light-touch AI approach at G20 technology talks
Tesla publishes European FSD safety dashboard before possible approval vote
OpenAI adds a Healthcare Public Data plugin for eligible US clinicians
India prepares a Unified Agent Protocol for UPI payments.
EU regulators ask publishers about Google AI-search opt-outs
Diraq and Equinix plan an eight-qubit silicon-spin computer in a Sydney data centre
Broadcom introduces AI-ready data foundations in VMware Tanzu Platform
Anthropic introduces Claude Fable 5.1 and restricted Mythos 5.1
