Australian AI news
When a story moves a verdict, the review gets a new dated edition and the change is logged on the review timeline.
Claude Max class action: what the US case means for Australians
A US suit says Claude Max gives far less than 5x or 20x Pro. Anthropic says the multiple is per session. Australians are not in the proposed class.
Opus 5, Fable 5.1 and Mythos 5.1 — what the Claude 5 family means for Australians
Anthropic's top tier is now three models and Australians can buy two. The 30-day retention rule binds API organisations, not Pro subscribers.
Sora's API is gone. Here's what Australians should use instead.
OpenAI switched off the last way into Sora on 24 September. If anything you made in it still matters, try the export today. Then pick one of the three alternatives we've actually scored for Australians.
Australia's app-store age checks are now law. Most of it already happened.
The deadline fell on 9 September. Apple switched it on in February — and the half that reaches AI tools has been live since March.
Flowise is shutting down: what Australian self-hosters should do now
Less than a year after Workday bought it, the open-source Flowise project is finished: code frozen, repository archived, core team gone on 31 August. Here's what an Australian self-hoster should do, and the three alternatives we've scored.
The TPB's AI guidance is final: your accountant is still on the hook, not the chatbot
Australia's tax regulator has finalised its rules on how tax and BAS agents use AI. The headline hasn't changed: the agent stays responsible, not the software. What's new is that it's settled — and that it reaches AI which acts on its own.
Australia's new privacy bill says an AI's guess about you counts as collecting your data
If an AI works something out about you, the draft says that counts as collecting it. Submissions close 18 September.
From 20 October, telling the Fair Work Commission you used AI isn't optional
From 20 October, drafting your Fair Work case with a chatbot carries three obligations — and costs orders if you ignore them.
Europe is now enforcing its AI law — what that changes for the tools Australians use
Since 2 August 2026, Brussels can enforce its AI rulebook against the labs behind ChatGPT, Claude, Gemini and Copilot. Australia didn't write the law — but Australians will feel it anyway.
Claude Code & Cowork: What Changed in 60 Days — and What It Means for Australians
Anthropic's two agent products had a busy two months. Here's what actually matters for Australian users.
AI in May 2026: the office tools grew agents — and meters
Notion 3.5, Atlassian Team '26, Linear Code Intelligence, ClickUp Brain² — May was when workspace AI matured, and the free agent trials began to meter.
AI in March 2026: Anthropic lands in Sydney, ChatGPT brings ads
A Sydney office for Anthropic, ads on ChatGPT in Australia, Claude inside Microsoft Copilot — March was the month the AU market got real attention, good and bad.