Orca Security’s Post

Why Your Password Storage is Fine, But Your File Checksum is Obsolete In the world of MLOps and complex development, you are constantly moving multi-gigabyte model artifacts and sensitive datasets. To verify that data integrity is maintained, you often see checksums using MD5 or SHA1. But wait, aren't those algorithms deprecated? The question boils down to a critical distinction: the difference between a cryptographic hash (for security) and a checksum (for integrity). Using the wrong tool can lead to silent data corruption or a security breach. This guide breaks down the MD5 vs. SHA-256 problem, explains when to use which tool, and tells you why we built a simple, client-side utility to manage data integrity without compromising security. The fundamental issue with MD5 (Message-Digest Algorithm 5, 128-bit) and SHA1 (Secure Hash Algorithm 1, 160-bit) is their vulnerability to collision attacks. A hash collision means two different inputs produce the exact same output hash. If a hash function is used for password storage or digital signatures, collision v https://lnkd.in/gpEys2tV

Well, we know security teams don't suffer from a lack of data nowadays. Alerts, vulns, logs, feeds… the pile never stops growing. So the bigger problem is what happens after collection: ▪️ A vuln scan says “critical.” ▪️ The SOC says “maybe.” ▪️ The app owner says “not in prod.” ▪️ Meanwhile, controls like WAF or EDR might already cover it… or not. That limbo is where time, energy, and trust get burned. The article below shows what it looks like when orchestration gives you the change needed to move from endless triage to validated, defensible actions that teams can actually execute. [link in comments]

The Model Context Protocol (MCP) is a standardized framework that bridges LLMs and enterprise systems with seamless automation. That bridge also expands the attack surface in ways defenders often overlook. In our latest Threat Labs blog, we explore two subtle, but devastating vectors: indirect prompt injection and RUG Pull attacks. Read the blog.

The Model Context Protocol (MCP) is a standardized framework that bridges LLMs and enterprise systems with seamless automation. That bridge also expands the attack surface in ways defenders often overlook. In our latest Threat Labs blog, we explore two subtle, but devastating vectors: indirect prompt injection and RUG Pull attacks. Read the blog.

The Model Context Protocol (MCP) is a standardized framework that bridges LLMs and enterprise systems with seamless automation. That bridge also expands the attack surface in ways defenders often overlook. In our latest Threat Labs blog, we explore two subtle, but devastating vectors: indirect prompt injection and RUG Pull attacks. Read the blog.

I believe MCP will offer both quality of life improvements for cyber professionals as advanced tools become more intuitive to engage with while also causing migraines as cyber professionals try to secure clients and tools servers integrating tools with AI. Rug pull in particular is interesting as systems like Docker MCP Gateway offer attractive ways to leverage the work of communities of developers working on MCP. Anyways, great article promoting the understanding of these attacks vectors and threats.

The Model Context Protocol (MCP) is a standardized framework that bridges LLMs and enterprise systems with seamless automation. That bridge also expands the attack surface in ways defenders often overlook. In our latest Threat Labs blog, we explore two subtle, but devastating vectors: indirect prompt injection and RUG Pull attacks. Read the blog.

The Model Context Protocol (MCP) is a standardized framework that bridges LLMs and enterprise systems with seamless automation. That bridge also expands the attack surface in ways defenders often overlook. In our latest Threat Labs blog, we explore two subtle, but devastating vectors: indirect prompt injection and RUG Pull attacks. Read the blog.