15 Oct 2019

AI development has major security, privacy and ethical blind spots

A recent survey by O’Reilly sheds light on common risk blind spots of artificial intelligence (AI)/Machine learning (ML) developers. The most glaring oversight is security, with nearly three in four (73%) respondents acknowledging that they do not test their models for security flaws during development. Other issues that are commonly

Read More
03 Sep 2019

China’s Red-Hot Face-Swapping App Provokes Privacy Concern

Once again, a popular mobile app for generating deepfakes, i.e. images or videos doctored by artificial intelligence (AI), has prompted a major privacy backlash. Since last weekend, Chinese face-swap app Zao has taken China’s iOS store by storm. The app can generate deepfakes of scenes from popular movies and TV-shows

Read More
02 Sep 2019

US Unleashes Military to Fight Fake News, Disinformation

The US Defense Advanced Research Projects Agency (DARPA) aims to develop software that can spot fake news stories as well as fake audio, images and video (deepfakes) in order to combat “large-scale, automated disinformation attacks.” Over a period of 48 months the DARPA initiative, called the Semantic Forensics (SemaFor) program,

Read More
30 Aug 2019

The Implementation Of Facial Recognition Can Be Risky. Here’s Why..

While facial recognition holds great promise for the security industry as a more secure authentication mechanism than passwords or PINS, the technology does come with various risks according to Allerin CEO Naveen Joshi. The risks stem from the fact that facial recognition algorithms, like other technologies leveraging artificial intelligence (AI),

Read More
07 Aug 2019

DARPA Is Taking On the Deepfake Problem

The United States Defense Advanced Research Projects Agency (DARPA) is launching an initiative to battle the growing threat of deepfakes, which are images or videos doctored by artificial intelligence (AI) that show individuals saying and doing things they never said or did. According to the agency, deepfakes are increasingly used

Read More
19 Jul 2019

Researchers Easily Trick Cylance’s AI-Based Antivirus Into Thinking Malware Is ‘Goodware’

Security researchers with Skylight Cyber have found a surprisingly easy way to let malware bypass Cylance’s AI-based anti-malware solution. The research shows that while artificial intelligence holds great potential for cybersecurity, AI-driven security offerings can be far from bulletproof.  In order to deceive Cylance’s algorithm, all the researchers had to

Read More
24 Jun 2019

Researchers develop a technique to vaccinate algorithms against adversarial attacks

Researchers with CSIRO’s Data61 have developed a method to employ machine learning in order to “vaccinate” systems against adversarial attacks, which are attempts to tamper with machine learning models by feeding them malicious data. For instance, by distorting images in various ways, threat actors may be able to bypass surveillance

Read More
28 May 2019

Volume and quality of training data are the largest barriers to applying machine learning

A new Alegion report highlights the difficulties companies encounter when trying to develop artificial intelligence (AI) and machine learning (ML) solutions. The survey found that almost 8 in 10 firms (78%) the have launched AI/ML projects saw their initiatives came to a grinding halt before deployment. In addition, virtually every

Read More
16 May 2019

The AI Supply Chain Runs on Ignorance

It is no secret that in order to improve the performance of artificial intelligence (AI) solutions, developers need to feed massive amounts of data into their AI algorithms so that these can evolve through “learning.” However, experts warn that AI firms are often far from transparent about how they gather

Read More
09 May 2019

Artificial Intelligence May Not ‘Hallucinate’ After All

A new MIT study suggests that artificial intelligence (AI) may sometimes misidentify objects not because it sees things that aren’t there, but because it sees the world in ways that humans cannot. In previous studies, researchers discovered that while machine learning algorithms can be used to train AI to classify

Read More