
Good morning.
If you run a small business, work on the tools, or just use AI to get through the day, today’s news is mostly about practical shifts: clearer safety rules for younger users, massive new investment in the computing power that runs these tools, and moves that could make advanced AI more available for Aussie firms.
In today’s briefing:
📱 OpenAI rolls out a dedicated ChatGPT for teens with study tools and stronger guardrails
🏗️Nvidia backs a huge new OpenAI data centre with tens of billions in support
🖥️Australia wants data-centre operators to give local startups and researchers access to computing power
⚠️Deepfake scams using the PM’s image are costing Australians millions
🛠️ Tool Spotlight Claude: Still one of the strongest everyday options
AUSTRALIA AI WATCH
🖥️Australia Wants Data-Centre Operators To Give Local Startups And Researchers Access To Computing Power
Assistant Minister Andrew Charlton outlined plans for data-centre operators (including the big global players) to make computing power available on favourable terms to Australian startups, researchers and innovators. The goal is to move Australia higher up the AI value chain rather than just hosting infrastructure.
Why it matters for you: More local access to serious computing power could lower costs and barriers for Aussie small businesses and technical teams building their own AI tools or testing ideas.
Sources: AFR | Startup Daily
⚠️Deepfake Scams Using The PM’s Image Are Costing Australians Millions

ASIC warned that deepfake videos of Anthony Albanese and other public figures have been used in investment scams that cost Australians $7.4 million. The videos look and sound real.
Why it matters for you: Never trust a celebrity or politician endorsement for an investment you see online. Check the actual company on ASIC’s register and treat any “guaranteed returns” claim as a red flag.
Sources: The Guardian | ABC News
TOP STORIES
📱 Openai Launches Chatgpt For Teens The Rundown

The Daily Rundown: OpenAI has started rolling out a special version of ChatGPT for users aged 13–17. It automatically applies tighter limits on sensitive topics, adds a Study Mode that guides students step-by-step instead of just giving answers, and includes parental controls for quiet hours and safety alerts. Full availability in Australia is expected early next month.
Practical takeaway: If you have teens using AI for schoolwork, this version is designed to reduce shortcuts and risky chats. Parents can link accounts for better oversight. For businesses that employ young workers or run training programs, it shows the direction major tools are heading on age-appropriate use.
Source: OpenAI
🏗️ Nvidia Puts Serious Money Behind Openai’s Next Data Centre
The Daily Rundown: Nvidia is investing and providing credit support of up to $105 billion for a massive new AI data centre campus in Ohio that will supply computing power to OpenAI. The project is one of the largest of its kind and highlights how hungry the industry still is for chips, power and space.
Practical takeaway: More computing capacity should eventually ease pressure on prices and availability of the big AI tools most of us use. In the short term it doesn’t change your monthly bill, but it shows the scale of investment still needed just to keep the current wave of AI running.
AI TOOL SPOTLIGHT
Claude

Claude (claude.ai) – Still one of the strongest everyday options for longer documents, careful reasoning and business writing. You can upload files, keep project context, and get clear, structured answers without the hype. Free tier available; paid plans start from around $20 USD/month. Useful for drafting quotes, rewriting client emails in plain English, or checking contracts.
💡PROMPTING TIP OF THE DAY
When you want AI to write something practical for work, tell it the exact audience and the length.
Ready-to-use example: “Write a short, friendly follow-up email to a client after a site visit. Keep it under 80 words. Mention we measured the kitchen and will send a fixed quote by Friday. No jargon.”
FROM READING TO DOING
The deepfake investment scams costing Australians $7.4 million are the sharpest version yet of "confident and false" — a video that looks and sounds real, from someone who never said it. How to Tell When AI Is Lying to You builds the verification reflex for exactly this: treating a confident tone as no evidence at all, and checking before trusting.
And today's prompting tip — naming the audience and the length before you ask — is the small habit that separates a usable draft from a generic one, which is the whole shift Talking to AI Like a Person (Not a Search Engine) teaches: brief it like a colleague, not a search box. Both are in the AI Foundations track at Vortex Academy.


