Security

Photon framework scales AI vulnerability discovery
Oak Ridge National Laboratory's Center for Artificial Intelligence Security Research (CAISER) is shining a light... Read more
Will a new border deal with the US open a backdoor into New Zealanders' personal data?
Anyone who has recently traveled to the United States will be familiar with biometric checks—facial... Read more
Hybrid AI tool unmasks hidden digital abuse to transform forensic investigations
Researchers from the University of Huddersfield have developed a hybrid AI tool to detect patterns... Read more
AI making cyber attacks costlier and more effective: Munich Re
Artificial intelligence is making cyberattacks increasingly sophisticated and costlier for businesses, reinsurer Munich Re said... Read more
New framework addresses privacy, dignity risks posed by modern AI systems
In a new article, researchers introduce the capabilities approach–contextual integrity (CA-CI), a framework that addresses... Read more
Study finds AI privacy leaks hinge on a few high-impact neural network weights
Researchers have discovered that some of the elements of AI neural networks that contribute to... Read more
Sensor chips help identify deepfakes by adding cryptographic signatures to camera data
AI-generated images and videos pose a threat to democratic processes and undermine trust within society.... Read more
Your smart home can be easily hacked. New safety standards will help, but stay vigilant
On a quiet suburban street, a modern Australian home wakes before its owners do.... Read more
'Neuron-freezing' technique can stop LLMs from giving users unsafe responses
Researchers have identified key components in large language models (LLMs) that play a critical role... Read more
Study: 'Security fatigue' may weaken digital defenses
From password resets and software updates to phishing alerts and cybersecurity trainings, today's workplace is... Read more
Study finds no single tool can fully protect online financial data
Research in the journal Electronic Government discusses the growing need for protecting one's personal financial... Read more
AI model trained on 14,000 Urdu news stories spots misinformation with 96% accuracy
A deep learning model trained on more than 14,000 Pakistani news articles can spot misinformation... Read more
From demons to mega behemoths: How 'monstrous' scam networks are growing
New research led by the University of Portsmouth uncovers how scammers operate worldwide, dividing them... Read more
Grid vibrations: AI detects power supply cyberattacks in less than two seconds
Modern energy infrastructure is increasingly defined as cyber-physical systems where physical power distribution and digital... Read more
Why harmful content keeps reaching children online, and what advertising has to do with it
Children today can encounter harmful material online with alarming ease, including violent, sexual and self-harm... Read more