Tag: arbitrary code execution
37 articles
AI Models Outpace GPT-5.5 in Chrome Vulnerability Exploits
Meet ExploitBench, a groundbreaking benchmark that puts AI models to the test, pushing them to go beyond mere vulnerability detection and actually exploit real-world flaws - and the results are in. This innovative tool, developed by Bugcrowd and Carnegie Mellon University experts, grades AI models on their ability to chain discoveries into usable exploits, revealing surprising capabilities.
Hackers Exploit FortiClient Flaw to Deliver Infostealer Malware
Hackers are exploiting a vulnerability in FortiClient Enterprise Management Server to deliver infostealer malware, cleverly disguising the payload as a legitimate Fortinet endpoint update. This sneaky tactic uses FortiClient-managed VPN scripting workflows to execute the malicious code, putting security teams on high alert.
LiteSpeed Plugin Flaw Exploited to Run Scripts as Root
A critical flaw in the LiteSpeed plugin, CVE-2026-48172, is being actively exploited to give cPanel users unlimited power, allowing them to run scripts as root. This severe vulnerability, rated 10.0 on the CVSS scale, puts your online security at risk and demands immediate attention.
Drupal Sites Targeted in SQL Injection Attacks
Drupal sites are under attack as SQL injection exploits are now being detected in the wild, taking advantage of a vulnerability that can be triggered without authentication. This critical flaw, CVE-2026-9082, allows attackers to execute arbitrary SQL and potentially run remote code, putting sites that use PostgreSQL at risk.
ChromaDB Flaw Enables Server Hijacking via AI Model Exploit
A newly discovered vulnerability, CVE-2026-45829, in ChromaDB's Python FastAPI variant allows hackers to hijack servers by exploiting AI models, with a security expert noting that authentication is present but poorly placed. This flaw lets unauthenticated attackers run arbitrary code on exposed servers by cleverly manipulating API endpoints.
Palo Alto Networks Exploits Critical PAN-OS Flaw in Limited Attacks
Palo Alto Networks has patched a critical flaw in its PAN-OS software, CVE-2026-0300, which allowed hackers to execute malicious code with root privileges - and the company says it's already been exploited in targeted attacks. The vulnerability, a buffer overflow in the User-ID Authentication Portal service, could be triggered by sending specially crafted packets.
SAP Patches Critical Flaws in Commerce Cloud and S/4HANA
SAP has patched a critical vulnerability in its Commerce Cloud and S/4HANA systems, warning that hackers could exploit the flaw to upload malicious code and take control of the application. This security gap, caused by a misconfigured Spring Security setup, put sensitive data and system integrity at risk.
cPanel Discloses Fixes for High-Severity Vulnerabilities
cPanel has patched three high-severity vulnerabilities, including a critical flaw that allows hackers to execute arbitrary Perl code, putting your online data at risk. The fixes address weaknesses that could be exploited to read sensitive files, disrupt service, or execute malicious code.
Palo Alto Firewalls Targeted in Active Exploitation
Thousands of Palo Alto firewalls are at risk due to an actively exploited vulnerability, CVE-2026-0300, that allows hackers to execute arbitrary code with root privileges. This alarming flaw affects 5,821 internet-exposed VM-Series firewalls, leaving them open to potential cyber attacks.
Palo Alto Networks Firewalls Targeted in Zero-Day Exploits
Palo Alto Networks firewalls are under attack by zero-day exploits targeting a vulnerability in the User-ID Authentication Portal, allowing hackers to execute malicious code with root privileges. This buffer overflow flaw, tracked as CVE-2026-0300, poses a significant risk to organizations with Internet-exposed firewalls.
Google Fixes Critical Gemini CLI Flaw Enabling Remote Code Execution
Google patched a critical flaw in Gemini CLI that allowed hackers to inject malicious code and take control of host systems, thanks to a report from Novee Security. The vulnerability, scoring a perfect 10.0 on the CVSS scale, has been fixed in recent updates to the @google/gemini-cli and google-github-actions/run-gemini-cli packages.
SGLang Flaw Enables Remote Code Execution via Malicious Model Files
A single malicious file can become a powerful gateway for attackers to run arbitrary commands on vulnerable machines - and a newly disclosed flaw in SGLang, CVE-2026-5760, reveals just how easily this can happen through specially crafted GGUF model files. This highly severe vulnerability, scoring 9.8 out of 10.0, enables remote code execution on systems that trust it.
Cisco Fixes Flaws Enabling Code Execution in Identity Services, Webex
Cisco has patched four critical vulnerabilities in its Identity Services and Webex Services, which could have allowed attackers to run arbitrary code and impersonate any user, posing a massive security risk. The fixes address flaws with CVSS scores as high as 9.8, safeguarding against devastating attacks.