Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

Security Operations Centers (SOCs) are the command center of an organization’s frontline cybersecurity defenses — responsible for monitoring threats, prioritizing alerts, and orchestrating remediation. However, today’s SOCs are facing an existential crisis: an overwhelming volume of increasingly complex and sophisticated threats combined with a shortage of skilled analysts.

There are rare moments when the technology world takes a massive leap forward. Today is one of those moments. At Salt Security, we have always believed that APIs form the connective tissue of the modern enterprise. As AI accelerates the pace of business beyond anything we’ve seen before, a simple truth has become clear: The way we discover, understand, and secure APIs must evolve, or we will be left behind.

MCP rapidly enhances AI capabilities but introduces security challenges through its distributed architecture. Especially, the distributed nature of MCP requires a lot of NHIs and their secrets. Our research shows that MCP is a new source of leaks that already discloses real-world secrets.

In this post, we introduce our “Adversarial AI Explainability” research, a term we use to describe the intersection of AI explainability and adversarial attacks on Large Language Models (LLMs). Much like using an MRI to understand how a human brain might be fooled, we aim to decipher how LLMs can be manipulated.

AI is transforming productivity across every industry—from marketing and design to legal and engineering. But while employees rush to embrace tools like ChatGPT, Gemini, and Microsoft Copilot, many are using other tools without oversight from IT or security. As this grassroots usage grows, so does the volume—and sensitivity—of data flowing into AI tools.