AI hallucinations, or confidently incorrect outputs, are creating security challenges by misleading human operators in critical infrastructure settings.
Shortly after its public disclosure, attackers began exploiting CVE-2026-44338, an authentication bypass vulnerability in PraisonAI that exposes sensitive endpoints.
Google has introduced Intrusion Logging as a new opt-in feature on Android devices to support forensic analysis of sophisticated spyware attacks. This capability is included within...
The US House Committee on Homeland Security requests Instructure executives to testify on two cyberattacks by the ShinyHunters group targeting the Canvas platform, resulting in stolen...
The TeamPCP group has compromised npm and PyPI packages from several prominent organizations including TanStack and Mistral AI, embedding malicious code as part of the Mini...
Security operations centers face challenges investigating critical alerts, with key high-risk categories frequently overlooked due to operational blind spots rather than sheer alert volume.
A new variant of the TrickMo Android banking trojan has been discovered using The Open Network (TON) for its command-and-control operations, actively targeting users in several...
A malicious repository on Hugging Face disguised as OpenAI's Privacy Filter open-weight model reached the platform's trending list and delivered a Rust-based information stealer targeting Windows...
German police have dismantled a rebooted version of the Crimenetwork marketplace, which reportedly generated over 3.6 million euros in illicit revenue, leading to the arrest of...
A critical vulnerability in Ollama allows remote attackers to leak process memory through an out-of-bounds read flaw, potentially affecting hundreds of thousands of servers.
May 14, 2026
