AI Model Hidden Backdoor Vulnerability
Changing AI behavior is much easier and cheaper than many people thought. Cyber expert Katie Paxton-Fear put a hidden flaw into an open AI model. Amazingly, she finished this task in just about one...
Changing AI behavior is much easier and cheaper than many people thought. Cyber expert Katie Paxton-Fear put a hidden flaw into an open AI model. Amazingly, she finished this task in just about one...
A dangerous vulnerability has been discovered in the Trust Wallet browser extension, potentially allowing attackers to steal users’ cryptocurrency. The issue affected version 2.68, and the wallet’s team officially urged everyone who had installed...
Hacker groups have exploited a security gap in Array AG Series corporate gateways, implanting covert management micro-programs and creating fraudulent user accounts without the administrator’s knowledge. The flaw, tracked as CVE-2025-66644, enabled arbitrary command...
In mid-autumn 2025, researchers at Cyble and Seqrite Labs observed a new wave of targeted malicious activity dubbed Operation SkyCloak. According to their findings, an unidentified threat actor has been conducting a phishing campaign...
For over a year, programmer Harishankar Narayanan had been using his iLife A11 robot vacuum without incident—until he noticed that the device was constantly transmitting data to servers in China. When he attempted to...
The malicious modification of Telegram X, discovered by specialists at Doctor Web, turned out to be far more than a simple espionage tool—it is a fully fledged platform for the covert takeover and remote...
Several extensions for Visual Studio Code, distributed via the open OpenVSX repository, have been found to contain malicious code aimed at cryptocurrency theft and the compromise of developers’ machines. The campaign is attributed to...
CERT-UA has detected a new targeted campaign against Ukraine in which operators distribute malicious XLL modules inside ZIP archives via the Signal messenger, ultimately delivering the CABINETRAT backdoor to victims’ machines. The agency attributes...
Zscaler ThreatLabz, in its recent report, disclosed the details of a newly observed malware family dubbed YiBackdoor, first detected in June 2025. From the outset, analysts noted substantial source-code commonalities with the IcedID and...
According to Google Threat Intelligence, the China-linked espionage group UNC5221 has since March conducted a series of successful intrusions into corporate networks, exploiting previously unknown vulnerabilities in Ivanti products. These operations resulted in the...
Washington and London have at last managed to reach an accord on an issue that threatened to escalate into a serious diplomatic and technological conflict. U.S. Director of National Intelligence Tulsi Gabbard announced that...
For nearly a year, a malicious module known as Plague evaded detection by Linux security solutions, despite its active proliferation and deep entrenchment within one of the system’s most critical components—the authentication stack. Its...