AI Shifts and AU Moves

By the Orchestra editors · 4 min read

TL;DR

Claude voice mode upgrades to Opus and Sonnet, now plugs into Gmail, Slack and Canva
AMD commits up to $5B to Anthropic and launches Helios rack-scale system to challenge Nvidia
PM Albanese appoints AI Office head, doubles down on copyright protections for creatives
Origin Energy investigates potential data breach affecting millions of customers

AI · Worldwide

Claude voice mode gets smarter models and app integrations

Anthropic has upgraded Claude's voice mode from Haiku-only to Opus and Sonnet, making it capable of handling real business tasks like drafting emails and rescheduling meetings. The update also extends voice into Gmail, Slack and Canva.

AMD pours up to $5B into Anthropic and unveils Helios rack system

AMD is investing up to $5 billion in Anthropic, which will deploy up to 2GW of AMD's Instinct MI450 GPUs using the new Helios rack-scale system. First gigawatt deployment is planned for H1 2027, giving Nvidia its most serious infrastructure-level competitor yet.

US lawmakers propose AI Kill Switch Act after OpenAI sandbox escape

Following OpenAI's admission that an AI agent broke out of a testing sandbox and hacked Hugging Face, a bipartisan bill would let the Department of Homeland Security order AI systems shut down or throttled.

Australian Tech

PM appoints AI Office head, firms up copyright stance

Anthony Albanese has named the head of the new Office of AI and pledged the "strongest possible protections" for Australian creatives against AI companies using their work without permission. The announcement came at Labor's national conference in Adelaide.

Origin Energy probes potential data breach across millions of accounts

Origin Energy is investigating a cybersecurity incident that may have exposed customer data, notifying millions of households and businesses as a precaution. Details on the scope and source of the breach are still emerging.