Risks and Vulnerabilities

Generative AI servers are leaking images, workflows, and prompts. Are they creating risk in your AI supply chain?

How are generative AI technologies leaking data? In this report we analyse prompt leaks from llama.cpp servers.

Learn how to detect CVE-2025-29927, a critical vulnerability in Next.js that allows attackers to easily bypass middleware authorization measures.

Explore how attackers are exploiting Ollama instances and learn what you can do to secure them across your infrastructure.

Misconfigured AI APIs can expose sensitive data and security risks. See what our research uncovered.

Learn how to detect SQL injection and path traversal vulnerabilities across your infrastructure.

Learn how to detect and mitigate CUPS vulnerabilities, including CVE-2024-47176.

Learn about two critical vulnerabilities affecting the ServiceNow platform (CVE-2024-4789 and CVE-2024-5217) and how UpGuard can help.

Learn the best course of action for preventing Remote Code Execution through an exploitation of regreSSHion.

Learn more about email messaging protocols and how to protect exposed ports.

With the continuing development of the Internet of Things come new attack surfaces and cybersecurity risk directly related to the IoT.

Safety and security are critical when using programmable logic controllers for industrial manufacturing and critical infrastructure.

Cybercriminals can chain vulnerabilities in ScreenConnect, leveraging authentication bypass first and then moving through the system with path traversal.

Dual vulnerabilities allow remote code execution, leaked credentials, and privilege escalation among Microsoft Outlook and Microsoft Exchange Server.

VPN internet connections protect web traffic through an encrypted tunnel, keeping sensitive data from interception. So what are the risks?

Git is widely used for version control of production code across industry sectors. Keep your repositories and your applications protected.