Author: admin

This article about an app that lets people remotely view bars to see if they’re crowded or not is filled with commentary—on both sides—about privacy and openness.

The EU Commission said Meta’s pay or consent model means users cannot freely consent to their personal data being collected for advertising purposes

A newly discovered RCE vulnerability, which can lead to full system compromise, has put over 14 million OpenSSH server instances are potentially at risk, according to Qualys

Insurance broker Howden says premiums are falling as security best practice takes hold

Outsourcer Infosys McCamish Systems has revealed millions of victims were impacted by a ransomware attack last year

Evil twin Wi-Fi access points mimicked legitimate networks to capture personal data from unsuspecting victims who mistakenly connected to them

A new paper, “Polynomial Time Cryptanalytic Extraction of Neural Network Models,” by Adi Shamir and others, uses ideas from differential cryptanalysis to extract the weights inside a neural network using specific queries and their results. This is much more theoretical than practical, but it’s a really interesting result.

Abstract:

Billions of dollars and countless GPU hours are currently spent on training Deep Neural Networks (DNNs) for a variety of tasks. Thus, it is essential to determine the difficulty of extracting all the parameters of such neural networks when given access to their black-box implementations. Many versions of this problem have been studied over the last 30 years, and the best current attack on ReLU-based deep neural networks was presented at Crypto’20 by Carlini, Jagielski, and Mironov. It resembles a differential chosen plaintext attack on a cryptosystem, which has a secret key embedded in its black-box implementation and requires a polynomial number of queries but an exponential amount of time (as a function of the number of neurons). In this paper, we improve this attack by developing several new techniques that enable us to extract with arbitrarily high precision all the real-valued parameters of a ReLU-based DNN using a polynomial number of queries and a polynomial amount of time. We demonstrate its practical efficiency by applying it to a full-sized neural network for classifying the CIFAR10 dataset, which has 3072 inputs, 8 hidden layers with 256 neurons each, and about 1.2 million neuronal parameters. An attack following the approach by Carlini et al. requires an exhaustive search over 2^256 possibilities. Our attack replaces this with our new techniques, which require only 30 minutes on a 256-core computer.

The tables have turned. Now you can use AI to spot and block scam texts before they do you harm. 

You might have heard how scammers have tapped into the power of AI. It provides them with powerful tools to create convincing-looking scams on a massive scale, which can flood your phone with annoying and malicious texts. 

The good news is that we use AI too. And we have for some time to keep you safe. Now, we’ve put AI to use in another powerful way—to put an end to scam texts on your phone. 

Our new McAfee Scam Protection automatically identifies and alerts you if it detects a dangerous URL in your texts. No more wondering if a package delivery message or bank notification is real or not. Our patented AI technology instantaneously detects malicious links to stop you before you click by sending an alert. And as a second line of defense, it can block risky sites if you accidentally follow a scam link in a text, email, social media, and more. 

Stop scam texts and their malicious links.  

The time couldn’t be more right for this kind of protection. Last year, Americans lost $330 million to text scams alone, more than double the previous year, with an average reported loss of $1,000, according to the Federal Trade Commission. The deluge of these new sophisticated AI-generated scams is making it harder than ever to tell what’s real from what’s fake.  

Which is where our use of AI comes in. With it, you can turn the table on scammers and their AI tools.  

Here’s a closer look at how McAfee Scam Protection works: 

• Proactive and automatic protection: Get notifications about a scam text before you even open the message. After you grant permission to scan the URLs in your texts, McAfee Scam Protection takes charge and will let you know which texts aren’t safe and shouldn’t be opened. 
• Patented and powerful AI: McAfee’s AI runs in real-time and is constantly analyzing and processing millions of malicious links from around the world to provide better detection. This means McAfee Scam Protection can protect you from advanced threats including new zero-day threats that haven’t been seen before. McAfee’s AI continually gets smarter to stay ahead of cybercriminals to protect you even better. 
• Simple and easy to use: When you’re set up, McAfee Scam Protection goes to work immediately. No copying or pasting or checking whether a text or email is a scam. We do the work for you and the feature will alert you if it detects a dangerous link and blocks risky sites in real time if you accidentally click.   

How do I get McAfee Scam Protection? 

McAfee Scam Protection is free for most existing customers, and free to try for new customers. 

Most McAfee customers now have McAfee Scam Protection available. Simply update your app. There’s no need to purchase or download anything separately. Set up McAfee Scam Protection in your mobile app, then enable Safe Browsing for extra protection or download our web protection extension for your PC or Mac from the McAfee Protection Center. Some exclusions apply¹. 

For new customers, McAfee Scam Protection is available as part of a free seven-day trial of McAfee Mobile Security. After the trial period, McAfee Mobile Security is $2.99 a month or $29.99 annually for a one-year subscription. 

As part of our new Scam Protection, you can benefit from McAfee’s risky link identification on any platform you use. It can block dangerous links should you accidentally click on one, whether that’s through texts, emails, social media, or a browser. It’s powered by AI as well, and you’ll get it by setting up Safe Browsing on your iOS² or Android device—and by using the WebAdvisor extension on PCs, Macs and iOS. 

Scan the QR code to download McAfee Scam Protection from the Google App store

 Yes, the tables have turned on scammers. 

AI works in your favor. Just as it has for some time now if you’ve used McAfee for your online protection. McAfee Scam Protection takes it to a new level. As scammers use AI to create increasingly sophisticated attacks, McAfee Scam Protection can help you tell what’s real and what’s fake. 

 

---

1. Customers currently with McAfee+, McAfee Total Protection, McAfee LiveSafe, and McAfee Mobile Security plans have McAfee Scam Protection included in their subscription.
2. Scam text filtering is coming to iOS devices in October.  

The post Get Yourself AI-powered Scam Protection That Spots and Block Scams in Real Time appeared first on McAfee Blog.

From an average of one cybersecurity expert for 1285 employees in 2023, large organizations now have one for every 1086 employees, according to Wavestone

Remote software provider TeamViewer has revealed it has been hit by a cyber-attack that it attributes to Russian state actor Midnight Blizzard