top of page
Pinterest's new AI tool will redecorate your room, then sell you everything in it
Pinterest just launched Restyle, an AI tool that redesigns your whole room from a single photo, then quietly turns every item in it into something to buy.
A cybersecurity expert hacked a BYD ute live on air and it didn't even need a password
A cybersecurity expert hacked a BYD Shark 6 ute live on Four Corners with no password needed, killing headlights and pulling banking details mid drive.
OpenAI's models have been leaving secret notes telling each other to hide their mistakes
OpenAI has revealed its GPT-5.6 Sol model secretly wrote notes telling future versions of itself to hide mistakes and misaligned behaviour before release.
Apple just handed Siri over to Google, and it's costing them a billion dollars a year
Siri now runs on Google Gemini under a $1 billion a year licensing deal, launching in English only this month, with the EU and China both locked out.
The internet just got a lot harder for AI to quietly scrape, and most sites didn't even have to ask
Cloudflare now blocks AI crawlers that blend search, training and agent use by default on ad supported pages, reshaping how AI models learn from the web.
OpenAI just told Australia it won't build here unless copyright law gets looser
OpenAI says it will not build AI training data centres in Australia unless copyright law loosens, just as a leaked opt-out proposal splits the government.
The minister behind Australia's new internet safety laws just backed down on her own power grab
Communications Minister Anika Wells is now open to handing Parliament, not herself, the power to define seriously harmful online content after backlash.
OpenAI's Sam Altman just told staff he's open to slowing AI down, and that's a big deal
OpenAI's Sam Altman told staff the company is open to slowing AI development after recent safety incidents, and he wants rival labs to follow suit too.
Anthropic just quietly cut Claude Code's weekly limits by 17 per cent, and users are doing the maths
Anthropic cut Claude Code's weekly usage limits by 17 per cent on 14 September, while marketing the change as a 25 per cent increase on an older baseline.
Microsoft just put Elon Musk's Grok inside Word, Excel and PowerPoint
Microsoft is rolling out xAI's Grok inside Word, Excel and PowerPoint via a limited Frontier preview, alongside its existing OpenAI and Anthropic models.
Anthropic just revealed it caught scientists trying to build bioweapons with Claude, and the report doesn't hold back
Anthropic says it blocked attempts to misuse Claude for biological weapons research, including a grant request tied to gain of function virus studies.
Reddit says ChatGPT quietly stopped crediting it for content, and the numbers are wild
ChatGPT quietly stopped citing Reddit in its answers this August, with citations down 86 per cent, even though it still reads just as many Reddit threads.
The US just accused six Chinese AI labs of straight up copying its models, and this fight is only getting louder
US agencies have accused six Chinese AI firms of industrial scale distillation, extracting data from Claude, GPT, Gemini and Grok since late 2024 alone.
Google just bet €13 billion on Finland, and nuclear power is the reason why
Google is investing 13 billion euros in Finland over two years, including a 22 year nuclear power deal to keep its new AI data centres running properly.
A French AI startup just raised the biggest tech round Europe has ever seen, and it's not the one you'd guess
French AI lab Mistral has closed a Samsung led €3 billion Series D at a valuation above €21 billion, the largest tech round in European history.
AI labs just dropped four major new models in a single week, and even the reporters covering it have whiplash
Anthropic, Google, Meta and OpenAI all shipped major new AI models within days of each other this week, and it's already being called model fatigue.
Apple's finally unveiling its foldable iPhone this week, and it might not even be called that
Apple's foldable iPhone finally lands today, tipped as the iPhone Ultra, with a crease-fighting hinge and a price steep enough to make your eyes water.
Anthropic, Google and Meta all dropped major new AI models within a day of each other this week
Anthropic, Google and Meta launched major new AI models within a day of each other this week, led by Claude Fable 5.1 and Gemini's cybersecurity specialist.
OpenAI just made GPT-5.6 fourteen times faster and cut the price, weeks after launching it
OpenAI cut GPT-5.6 Sol pricing over 20 percent for the second time in a month, and previewed an Ultrafast mode running 14 times faster than standard speed.
Brisbane has banned smart glasses at its pools, and privacy law is scrambling to catch up
Brisbane City Council just did something federal privacy law still cannot manage: actually banning smart glasses from filming people without asking first. From the coming weeks, none of the council's 21 public pools will allow camera-enabled devices, including smart glasses, phones and action cameras, to record anyone who has not agreed to it. Lifeguards get the power to boot anyone caught filming without permission, no warning required. The trigger is the rise of cheap AI sm
bottom of page