Teslas are incredibly data-hungry, storing massive troves of data about their owners, including videos of crashes, location history, contacts and calendar entries from paired phones, photos of the driver and passengers taken with interior cameras, and other data; this data is stored without encryption, and it is not always clear when Teslas are gathering data, and the only way to comprehensively switch off data-gathering also de-activates over-the-air software updates for the cars, which have historically shipped with limited or buggy features that needed the over-the-air updates to fix them. Read the rest

Since 2016, there have been multiple instances of attacks on keyless entry car-locks, and there's a burgeoning industry of expensive ($5000) aftermarket alarm systems that are billed as protecting your car from these radio attacks on its security. Read the rest

Lime scooters have been recalled in Switzerland and cleared off the streets of New Zealand following a string of injuries, including multiple broken bones, caused by a software bug that brings the scooters to an abrupt halt, throwing their riders off (the scooters are still available in the USA despite an account of a similar incident in Texas). Read the rest

Last year, Apple outraged independent technicians when they updated the Iphone design to prevent third party repair, adding a "feature" that allowed handsets to detect when their screens had been swapped (even when they'd been swapped for an original, Apple-manufactured screen) and refuse to function until they got an official Apple unlock code. Read the rest

Researchers from the University of Toronto's Citizen Lab (previously) have published an extensive report on the image filtering systems used by Chinese messaging giant Wechat to prevent the posting of banned political messages and other "sensitive" topics that are censored in China. Read the rest

Last week, I linked to a critique of Google's new "confidential mode" for Gmail and Google Docs, which purports to allow you to send people documents without letting them print, copy or forward them. Read the rest

In Partial Information Attacks on Real-world AI, a group of MIT computer science researchers report on their continuing work fooling Google's image-classifier, this time without any knowledge of how the classifier works. Read the rest

Machine-learning-based image classifiers are vulnerable to "adversarial preturbations" where small, seemingly innocuous modifications to images (including very trivial ones) can totally confound them. Read the rest

Three researchers from Kyushu University have published a paper describing a means of reliably fooling AI-based image classifiers with a single well-placed pixel. Read the rest

An "adversarial perturbation" is a change to a physical object that is deliberately designed to fool a machine-learning system into mistaking it for something else. Read the rest

The Open AI researchers were intrigued by a claim that self-driving cars would be intrinsically hard to fool (tricking them into sudden braking maneuvers, say), because "they capture images from multiple scales, angles, perspectives, and the like." Read the rest

Netflix has become one of the main forces for DRM in the world, a driver behind the W3C's dangerous, web-scale DRM project, and now they've announced that their app will no longer run on rooted/bootloader unlocked Android devices, because these devices can run code that overrides Google Widevine DRM (Widevine doesn't work well under the best of circumstances, and it harbored unpatched, showstopper bugs since its very inception). Read the rest

In a newly revised paper in Computer Vision and Pattern Recognition, a group of French and Swiss computer science researchers show that "a very small perturbation vector that causes natural images to be misclassified with high probability" -- that is, a minor image transformation can beat machine learning systems nearly every time. Read the rest