Your new post is loading...
Your new post is loading...
AI systems have gotten more capable and easier to use
Les LLM sont vulnérables à un risque particulièrement vicieux : les agents dormants, une forme de comportement qui trompe la plupart des mesures de sécurité.
The US military aims to maintain its dominance by building autonomous attack drones that collaborate with humans and overwhelm defenses in swarms.
Wi-Fi-based Positioning Systems (WPSes) are used by modern mobile devices to learn their position using nearby Wi-Fi access points as landmarks. In this work, we show that Apple's WPS can be abused to create a privacy threat on a global scale. We present an attack that allows an unprivileged attacker to amass a worldwide snapshot of Wi-Fi BSSID geolocations in only a matter of days. Our attack makes few assumptions, merely exploiting the fact that there are relatively few dense regions of allocated MAC address space. Applying this technique over the course of a year, we learned the precise locations of over 2 billion BSSIDs around the world. The privacy implications of such massive datasets become more stark when taken longitudinally, allowing the attacker to track devices' movements. While most Wi-Fi access points do not move for long periods of time, many devices -- like compact travel routers -- are specifically designed to be mobile. We present several case studies that demonstrate the types of attacks on privacy that Apple's WPS enables: We track devices moving in and out of war zones (specifically Ukraine and Gaza), the effects of natural disasters (specifically the fires in Maui), and the possibility of targeted individual tracking by proxy -- all by remotely geolocating wireless access points. We provide recommendations to WPS operators and Wi-Fi access point manufacturers to enhance the privacy of hundreds of millions of users worldwide. Finally, we detail our efforts at responsibly disclosing this privacy vulnerability, and outline some mitigations that Apple and Wi-Fi access point manufacturers have implemented both independently and as a result of our work.
Avec l’accélération de la dématérialisation et la suppression de postes aux Finances publiques, la CGT affirme que le nombre d’erreurs causées par l’intelligence artificielle est en hausse, provoquant la colère des contribuables des Bouches-du-Rhône.
We have identified how millions of concepts are represented inside Claude Sonnet, one of our deployed large language models. This is the first ever detailed look inside a modern, production-grade large language model.
Of course Apple is designing its own AI chips for its data centers. This will give them an end-to-end, cloud-to-edge, AI system that its competitors will not be able to match any time soon.
Researchers are learning that handwriting engages the brain in ways typing can't match, raising questions about the costs of ditching this age-old practice, especially for kids.
OpenAI's Preferred Publisher Program offers media companies licensing deals.
Une étude compare la sécurité et le respect de la vie privée entre Android et iOS. Malgré une activité réseau plus intense, l'iPhone se montre plus prudent dans ses communications avec des serveurs étrangers potentiellement à risque, contrairement à son rival qui multiplie les échanges douteux.
Sources suspect China is behind the targeted exploitation of two zero-day vulnerabilities in Cisco’s security appliances.
L'infrastructure à mettre en place diffère grandement selon les cas d'usage. Tour d'horizon des configurations répondant à chacun.
This is our tenth annual landscape and “state of the union” of the data, analytics, machine learning and AI ecosystem.
In 10+ years covering the space, things have never been as exciting and promising as they are today. All trends and subtrends we described over the years are coalescing: data ha
|
AI models are always surprising us, not just in what they can do, but also in what they can't, and why. An interesting new behavior is both superficial
Use this Google Sheets Apps Script to find k-Nearest Neighbors across your site with a highly relevant cosine similarity score.
Les IA génératives, exportés dans le monde entier vont-elles engendrer une uniformisation des valeurs culturelles ? C’est ce que révélé une étude de 2022. Qu’en est-il aujourd’hui ?
L'Institut Paris Region réalise des études, enquêtes et recherches ayant pour objet l'aménagement et l'urbanisme dans la région Île-de-France. Acteur majeur de l'aménagement métropolitain durable et du devenir de la région, il peut prêter son expertise en France et à l'étranger.
EXCLUSIVE: Ultra-wideband radio has been heralded as the solution for “relay attacks” that are used to steal cars in seconds. But researchers found Teslas equipped with it are as vulnerable as ever.
Note: In my last newsletter, I said that my next post would be the second part of my Facebook autopsy. Don’t worry, that’s still coming, but given the recent drama between Sam Altman, OpenAI, and Scarlett Johansson, I felt the need to write something. Don’t worry, I
Human intelligence shaped organisations but it is no longer the only form at work
That pesky 'previously unknown software bug' strikes again
Quadrupeds being reviewed have automatic targeting systems but require human oversight to fire.
Meta's Twitter/X rival Threads is growing at a stable pace. The social network now has more than 150 million monthly active users.
What are neurorights? Elon Musk and Mark Zuckerberg are building tech to read our minds. Lawyers are trying to protect our mental privacy from Silicon Valley through neurorights.
Une étude de l'UNESCO révèle les biais de genre choquants présents dans les modèles de langage d'IA les plus populaires. Des stéréotypes néfastes qui soulèvent de sérieuses questions éthiques et appellent à une refonte en profondeur du développement de l'IA.
|