Red Canary Expands AI Innovations to Cut Alert Overload

Positioning security leaders as more than risk managers turns them into business enablers, trusted advisers, and, eventually, integral members of the C-suite.

Cybersecurity researchers have uncovered a jailbreak technique to bypass ethical guardrails erected by OpenAI in its latest large language model (LLM) GPT-5 and produce illicit instructions. Generative artificial intelligence (AI) security platform NeuralTrust said it combined a known technique called Echo Chamber with narrative-driven steering to trick the model into producing undesirable

Two known threat activity clusters codenamed Head Mare and Twelve have likely joined forces to target Russian entities, new findings from Kaspersky reveal. “Head Mare relied heavily on tools previously associated with Twelve. Additionally, Head Mare attacks utilized command-and-control (C2) servers exclusively linked to Twelve prior to these incidents,” the company said. “This suggests

Cybersecurity researchers have flagged a credit card stealing malware campaign that has been observed targeting e-commerce sites running Magento by disguising the malicious content within image tags in HTML code in order to stay under the radar. MageCart is the name given to a malware that’s capable of stealing sensitive payment information from online shopping…

The phishing operation is using Telegram groups to sell a phishing-as-a-service kit with customer service, a mascot, and infrastructure that requires little technical knowledge to install.

Hackers have long used Word and Excel documents as delivery vehicles for malware, and in 2025, these tricks are far from outdated. From phishing schemes to zero-click exploits, malicious Office files are still one of the easiest ways into a victim’s system. Here are the top three Microsoft Office-based exploits still making the rounds this…