A Russian-speaking threat actor successfully jailbroke Google’s Gemini AI to conduct sophisticated social engineering attacks that resulted in cryptocurrency wallet theft. The attacker leveraged the compromised AI to craft convincing phishing communications, targeting at least one victim with politi
Gunra ransomware evolved from Conti roots into a full-scale RaaS operation, hitting dozens of orgs globally in under a year with leak sites and affiliate programs
The U.S. Cybersecurity and Infrastructure Security Agency (CISA) has added a critical vulnerability in the LiteSpeed Cache plugin for WordPress to its Known Exploited Vulnerabilities (KEV) catalog. Tracked as CVE-2024-28000, this privilege escalation flaw affects millions of WordPress sites and is b
CRITICAL: 200,000+ WordPress sites exposed to full account takeover via Burst Statistics plugin flaw. Authentication bypass vulnerability now being exploited in the wild.
Threat actors are exploiting F5 BIG-IP appliances as entry points to compromise enterprise Linux networks. Attackers gain initial access through vulnerable BIG-IP devices, establish SSH backdoors, and pivot laterally into connected Linux infrastructure. Organizations using F5 BIG-IP in their network
Zimperium zLabs uncovered a 10-month Android malware campaign involving 250 fake apps that silently enrolled victims in premium SMS services across four countries. The operation abused Google’s SMS Retriever API and hidden WebViews to bypass user consent and intercept authentication codes.
A sophisticated supply chain attack dubbed “Miasma” has compromised multiple Red Hat-associated npm packages, injecting malicious code designed to steal credentials and propagate as a worm. The attack targeted the npm ecosystem, affecting downstream applications that depend on these packages. Organi
An author’s reliance on AI language models resulted in fabricated quotes being published in a book about truth and authenticity. Despite discovering the synthetic content, the author defended continued AI usage, highlighting a critical vulnerability in AI-assisted content creation: hallucination wit
A dangerous new class of AI attacks called prompt injection is targeting Kubernetes-hosted LLM applications in ways that traditional security tools cannot detect. CrowdStrike’s Falcon AIDR now delivers runtime visibility at the prompt layer, identifying malicious intent inside natural language interactions without proxies or architectural changes.
GitLab has released security updates addressing multiple critical vulnerabilities in both Community Edition (CE) and Enterprise Edition (EE), including severe flaws in its Duo AI features, authorization bypass issues, and denial-of-service (DoS) vectors. The most severe vulnerabilities allow unautho