The Latest

Binance’s $6.3 Billion Hack Began With a Fake Job Interview: What the Post-Mortem Says, and How to Spot a Fake Recruiter

An empty office chair facing a white desk with an open laptop, a coffee cup and a blank sheet of paper in a bright room
Binance’s initial findings trace the $6.34 billion theft of 29 October to a fake job interview in September: a fake HR call, a fake CEO, and a coding “assignment” that installed new malware on an engineer’s work computer. An internal AI tool then served as a springboard to two wallet engineers and a deployment token. What the report says, what it leaves out, who the FBI blames, and how companies and job seekers can protect themselves.
View Post

Set to 110 °C: The Night Hackers Reached a Blood Bank’s Plasma Freezers — and Why the Alarm Won

A calm, brightly lit medical cold storage room with a row of frosted stainless steel freezer doors and a small glowing blue wall panel
During the night of 1 to 2 November 2026, someone reached a refrigeration contractor’s remote supervision system and raised the settings of plasma cold rooms at three sites of France’s blood service EFS, reportedly to 110 °C. It was a setpoint, not a temperature. An alarm and an on-call technician caught it, and no blood products were lost. What is confirmed, what the viral posts get wrong, the unverified claim of responsibility, a clearly labelled hybrid-threat scenario, and what businesses and consumers should do.
View Post

We Gave 42 AI Setups the Same 15 Legal Questions. The Model Built for Law Scored 8 Out of 100

A brass balance scale on a bright white desk holding a small cobalt glass cube in one pan and a turquoise one in the other, beside a laptop and a stack of blank paper files
On Lexassay, my legal AI test, 42 AI setups answered the same 15 questions on German, US and EU law without internet. Z.ai’s GLM-5.3 FlashX Legal, built for law, scored 8 of 100; Europe’s Mistral Large 4 scored 10; Claude Opus 5.5 led with 79. Another test ranks them almost the other way round. What a wrong legal answer can cost, and how to protect yourself. Disclosure: Thorsten Meyer AI runs Lexassay; Claude agents grade. Not legal advice.
View Post

The Malware Came From the Factory: Google Takes Down a 1.3-Million-Device TV Box Botnet

A small plain black streaming box on a white sideboard next to a switched-off TV, with a thin turquoise light thread rising from the box
Google says the Chinese device maker Videostrong built a backdoor into the firmware of its Android TV boxes and sold access to criminals. In late October, Google and partners cut the XCPE botnet of more than 1.3 million boxes off from its servers. How the numbers were counted, why some infected boxes were certified, what is still unknown, and how to check the box under your TV.
View Post

An EU Court Adviser Says AI Chatbots Can Infringe Publishers’ Rights. Facts Stay Free, the Writing Does Not. What It Means for Companies, Publishers and Users

A stack of blank folded paper on a bright white desk next to a smartphone showing a glowing turquoise speech bubble, with a small brass balance scale by the window
In the first generative AI case at the EU Court of Justice, Advocate General Szpunar says chatbots that fetch news articles live and reuse them can infringe press publishers’ rights, but only if they get around access barriers or reproduce the article’s wording so recognisably that readers can skip the original. Facts stay free. On the actual dolphin article in the case, his test points against the publisher. What it means for companies, publishers and users. Not legal advice.
View Post