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

Robots can learn two-handed tasks from people’s handheld recordings alone, a Japanese group’s open model suggests
Japan’s AI Robot Association has released AIRoA-0, a robot model trained on 16,151 hours of people using a handheld gripper, which it says beat…
-

Prosecutors say AI-faked bosses and lawyers tricked companies out of $129m
US prosecutors have charged three men with using AI to fake the voices and faces of chief executives and law-firm lawyers to get staff…
-

An AI system found 948 Linux kernel bugs that hand attackers full control, OpenAI says
OpenAI says an upcoming AI model found and exploited 948 Linux kernel bugs in one run, and fixes for many are still on their…
-

India warns banks and fintechs that criminals are using AI agents to drain payment systems at machine speed
India’s cybercrime centre says AI agents are being used to break into financial firms’ payment systems, and sets out the controls it wants in…
-

AI is likely helping a frontier lab improve its models about twice as fast, outside review finds
An independent review inside Anthropic estimates its models are improving about twice as fast as before, mostly because AI is helping with the research.
-
Gemini Notebook: what it is good at, and what it is not
Learn what Google’s Gemini Notebook (formerly NotebookLM) does well, where it falls short, its free plan and the privacy settings to know.
-
How to learn a new topic with an AI tutor
Use an AI assistant as a tutor that explains, quizzes and checks your understanding, then verify what it teaches.
-
Malware now asks AI to pick its victims: what it means for you
Learn what the VOLTA infostealer does with AI and the basic steps that keep your most important accounts safe.
-
AI watermarking explained in 2 minutes
An AI watermark is a hidden signal woven into generated text, images, audio or video so software can later tell that AI made it.
-
AI agent monitoring explained in 2 minutes
AI labs now watch what their agents do, with layers of automated checks and human review, much like a bank watches for fraud.
