Collaborating Authors


Cybersecurity Tools Gaining an Edge from AI


In 2021, more firms will employ AI to battle cyberattacks, trying to gain an edge in a game of one-upmanship with hackers and attackers. A survey of 20 cybersecurity experts recently surveyed by Forbes showed some patterns. For example, open source software can be an easy way into organizations. Gaining more visibility into open source contributions is possible with the use of AI and machine learning, according to Maty Siman, CTO of Checkmarx, a software security company based in Ramat Gan, Israel. "Rarely does a week go by without the discovery of malicious open source packages," Siman stated.

What do AI, blockchain and GDPR mean for cybersecurity?


It's possible that someone may be watching your screen--by listening to it. A recent study from cybersecurity analysts at the universities of Michigan, Pennsylvania and Tel Aviv found that LCD screens "leak" a frequency that can be processed by artificial intelligence to provide a hacker insight into what's on a screen. "Displays are built to show visuals, not emit sound," says Roei Schuster, a PhD candidate at Tel Aviv University and a co-author of the study with doctoral candidates Daniel Genkin, Eran Tromer and Mihir Pattani. Yet the team's study shows that's not the case. The researchers were able to collect the noise through either a built-in or nearby microphone or remotely over Google Hangouts, for example.

A UK startup that uses AI to help you with GDPR has won $1m in funding


A UK startup that uses artificial intelligence to help businesses manage data and meet GDPR regulations has won $1 million in funding from a group of companies including Microsoft. Hazy has created a more secure data sharing system that lets people track and manage who has access to information, and generates GDPR compliant legal agreements. The company has won Microsoft's Innovate.AI global startup prize, which is open to small firms using AI to solve problems and improve lives. Hazy took home the Europe prize, along with up to $500,000 in Azure credits, Office 365 licences and $1 million in funding from M12, Microsoft's venture fund, and Notion Capital. Harry Keen, Chief Executive of Hazy, wrote in a blog post: "It's an amazing accolade for the Hazy team, which I'm proud to say is made up of some of the world's best AI and machine learning experts, thanks to our partnership with UCL.

MinerEye launches AI-powered Data Tracker to bolster GDPR compliance


With the May 25 deadline for the EU's General Data Protection Regulation (GDPR) fast approaching, a new tool from Israel-based MinerEye uses artificial intelligence (AI) to help identify all the data in your organization that needs protection under the new rules. Ahead of the RSA conference in San Francisco, MinerEye has launched its Data Tracker solution, which automates the process for detecting, tracking, and securing sensitive assets. According to a MinerEye press release, it can be used with unstructured and dark data, and can be leveraged as part of compliant cloud migration. "Companies cannot protect, manage or utilize information they can't find," MinerEye CEO and co-founder Yaniv Avidan said in the release. "Using our Interpretive AITM, MinerEye fuses computer vision and machine learning to track information at the byte and pixel level, which no other solution has achieved."

DeepSign: Deep Learning for Automatic Malware Signature Generation and Classification Machine Learning

This paper presents a novel deep learning based method for automatic malware signature generation and classification. The method uses a deep belief network (DBN), implemented with a deep stack of denoising autoencoders, generating an invariant compact representation of the malware behavior. While conventional signature and token based methods for malware detection do not detect a majority of new variants for existing malware, the results presented in this paper show that signatures generated by the DBN allow for an accurate classification of new malware variants. Using a dataset containing hundreds of variants for several major malware families, our method achieves 98.6% classification accuracy using the signatures generated by the DBN. The presented method is completely agnostic to the type of malware behavior that is logged (e.g., API calls and their parameters, registry entries, websites and ports accessed, etc.), and can use any raw input from a sandbox to successfully train the deep neural network which is used to generate malware signatures.