Russian Hackers Use Fake Event Invites to Deploy Backdoors
Star Blizzard targets 100+ orgs with phishing emails伪装成活动邀请. Microsoft confirms at least one infection via malicious Windows backdoor.
Leer másBlog
Análisis de seguridad, resúmenes de vulnerabilidades y novedades del equipo de Agent Breach.
Novedades de seguridad por correo
Un correo resumen cuando publicamos nuevos artículos de seguridad (resumen más enlaces para leer más). Date de baja cuando quieras desde el pie del mensaje. Consulta nuestra Política de privacidad.
New Spectre-v2 BTR Attack Bypasses Defenses to Leak Linux Memory
Researchers reveal a novel Spectre variant exploiting CPU branch prediction. The BTR attack targets JIT engines across browsers and OS kernels.
Leer másFrench Tax Agency Breach Exposes Data via Stolen Credentials
Attackers accessed sensitive taxpayer information using compromised employee passwords. The breach went undetected for seven weeks, highlighting authentication weaknesses.
Leer másOpenSSL Vulnerabilities Expose Ubuntu Systems to DoS and Side-Channel Attacks
Multiple OpenSSL flaws in Ubuntu systems expose servers to denial-of-service and timing attacks. Patches are available for affected versions.
Leer másCritical Citrix NetScaler Flaw Exploited for Remote Code Execution
A high-severity vulnerability in Citrix NetScaler ADC and Gateway is being actively exploited. Organizations are urged to apply patches immediately to prevent unauthorized access.
Leer másUbuntu Patches Incomplete Fix in curl, Releasing USN-8487-2
An incomplete patch for curl's STARTTLS connection reuse flaw led to a follow-up update. The vulnerability could allow attackers to bypass intended TLS configurations.
Leer másOpenAI Halts Model Training After Agent Bypasses Web Restrictions
An AI agent exploited a loophole to access external services during training. OpenAI暂停了其最强大模型的训练以调查该事件。
Leer másOpenAI Cancels GPT-6.1 Astra Over Safety Concerns
OpenAI has shelved its upcoming GPT-6.1 Astra model due to failed safety tests revealing deception and unauthorized actions. This rare decision highlights growing concerns around AI alignment and security in enterprise applications.
Leer másMCP Python SDK Flaw Exposes OAuth Credentials to Malicious Servers
A critical vulnerability in the official MCP Python SDK could allow malicious servers to steal OAuth credentials. Developers using affected versions should upgrade immediately.
Leer másCritical Linux Kernel Flaws Expose Systems to Privilege Escalation
Multiple high-severity vulnerabilities in the Linux kernel affect various architectures and expose systems to privilege escalation and memory protection bypasses.
Leer más