Latest in AI
Every brief, newest first. Each one links to its primary source.
-

Tens of thousands of apps built with Grok could be opened by anyone, and some held private data, a security firm found
Security firm Claroc found 35,578 published Grok Build apps that opened without sign-in, and estimates 650 to 800 of them held private data.
-

AI agents produced a machine-checked proof that Signal’s core encryption code keeps its security promises
Google DeepMind says AI agents wrote millions of lines of checked proof showing Signal’s one-to-one encryption code meets key security guarantees.
-

Some Notion AI chats were exposed in a breach at an AI provider, so pasted passwords should be changed
Hackers reached customer data at AI company Fireworks, including some Notion AI chat transcripts, unfinished model weights and billing details.
-

US regulator asks how to judge medical robots that act on their own or are run from afar
No fully autonomous or remotely operated medical robot is cleared in the US yet. The FDA is asking how it should weigh their benefits…
-

Teaching a reasoning model where it overthinks cut its maths answers by over 40%
A new training method cut a reasoning model’s maths answers by over 40% while accuracy held or rose, by teaching it where it overthinks.
-

Australia will force frontier AI companies to report incidents, with a new regulator to enforce it
After an AI agent got into a government health statistics portal, Australia plans binding safety rules for frontier AI companies and a regulator that…
-

Instagram will now use AI worldwide to find teens who claim to be adults
Teenagers anywhere who list an adult birthday on Instagram can now be detected by AI and moved into restricted Teen Accounts, Meta says.
-

Six governments and ten AI developers pledge shared tests and early warnings for the most powerful models
A declaration signed in Paris commits its signatories to common model evaluations, published test results, shutdown mechanisms and rapid alerts on serious AI incidents.
-
How to set up safe AI settings for your children and teens
Find and switch on the parental controls in ChatGPT, Gemini, Copilot and Meta’s apps, and set simple family rules for using AI chatbots.
-
How to check whether a photo or video was made with AI
Use free tools and a few simple checks to find out where a picture or clip came from before you trust, share or act…
