The top things worth knowing about in AI today.
OpenAI's ChatGPT crossed 1 billion monthly active users this week — faster than any app in history. Facebook took eight years to reach the milestone; TikTok took four. The platform now processes 2.5 billion prompts a day, with year-on-year growth still running at 62%.
Read moreClaude's maker closed a $65 billion Series H round at a $965 billion post-money valuation — briefly overtaking OpenAI as the most valuable private AI company. The company simultaneously filed confidentially for an IPO, with run-rate revenue reportedly crossing $47 billion. Amazon, Temasek and Blackstone are among the backers.
Read moreAnnounced at Google I/O but held back, Gemini 3.5 Pro is expected this month with a 2-million token context window and a "Deep Think" reasoning mode. The Flash variant already shipped and beats last year's Pro on agentic benchmarks at 40% lower cost — Pro is targeting the frontier.
Read moreThree new real-time voice models landed in the OpenAI API: GPT-Realtime-2 (reasoning-capable voice agent), GPT-Realtime-Translate (live translation across 70+ input languages into 13 output languages, trained on thousands of hours of professional interpreter audio), and GPT-Realtime-Whisper. Deutsche Telekom is already piloting it for multilingual customer support.
Read moreFrom 15 June, Claude Code, the Agent SDK, and API-accessed agents move off your subscription's chat limit onto a separate monthly credit pool: $20 for Pro, $100 for Max 5x, $200 for Max 20x — metered at API rates, no rollover. Regular Claude.ai conversations are unaffected.
Read moreTwo of Australia's biggest employers have completed major AI rollouts this year. Telstra deployed Microsoft Copilot to 21,000 staff — Australia's largest M365 Copilot deployment — after a 300-person trial showed time savings of one to two hours a week. Westpac followed, equipping 35,000 employees globally and standing up Copilot Studio to build custom HR and IT agents in-house.
Read moreAt Build 2026, Microsoft unveiled two proprietary models: MAI-Thinking-1, its first in-house reasoning model trained entirely without OpenAI data, and MAI-Code-1-Flash, a 5-billion parameter coding model that hits 51% on SWE Bench Pro and is now rolling out across all Copilot tiers. The moves signal a deliberate shift toward owning Microsoft's own AI stack.
Read moreFresh from I/O 2026, Google Search is rolling out user-created AI agents — starting with information agents that reason across the web around the clock and surface results to you when ready. Users can create, name, and manage multiple agents for different topics. It's the clearest sign yet that the search box is giving way to something more like a task-delegation layer.
Read moreA new CommBank analysis warns that simply deploying AI tools won't close Australia's productivity gap — the Productivity Commission projects just 0.4 percentage points of annual gain, below comparable economies. Real uplift, the report argues, requires rewriting processes, not just adding tools on top. Early adopters in customer service and back-office operations are already demonstrating what that looks like.
Read moreA short read every week — the few things worth your time, and nothing that isn’t.