PUBLIC DEFENDER By Brian Livingston The maker of the popular Claude large language model (LLM) — which became the number-one download from US app stores in February 2026 — recently announced a powerful service called Claude Mythos. The new LLM has reportedly discovered thousands of security holes in every major operating system and Web browser. […]

AI By Michael A. Covington Malfunctioning AI can hurt people. Preventing harm isn’t easy. To detect automatically whether an AI system is giving people bad advice, for instance, you’d need an AI that is smarter than the original one; and if you had it, you’d have used it in the first place. You see the […]

Whenever I hear or see headlines about a “supply chain vulnerability,” I think in terms of big enterprises and delivery of goods and services. You know — supplies. But in technology, the term “supply chain attack” means that an attacker has gotten into something trusted — such as a software vendor, service provider, hardware supplier […]

AI By Michael A. Covington Can we even measure it? That’s a hot question now. An article by E.K. Chen and others in the world’s top science journal, Nature, says AI has achieved human-level intelligence. We need to measure AI intelligence not only to test that claim, but also to tell whether one AI is […]

ISSUE 23.03 • 2026-01-19 AI By Michael A. Covington It’s easier than you might think to run LLMs (large language models) locally on your own PC, without connecting to a server. There are three reasons you might want to do so: to keep your data private, to avoid costs, and to avoid depending on commercial […]

Yesterday I was listening to a news show that focuses on economics. The discussion was about the PE ratio being the highest it’s been since the stock market crash and the dot-com bubble. A lot of the current excitement is about the promise of AI. And right now, there’s a lot of promise. In my […]

ISSUE 22.45 • 2025-11-10 Look for our BONUS issue on November 17, 2025!! AI By Michael A. Covington Yes, I mean 2025. We all know there was a paradigm shift in 2022–23, when AI suddenly came to mean generative LLM chatbots rather than machine learning or rule-based deduction, and people’s idea of what computers could […]

In Part 2, we’re diving headfirst into one of the most critical attack surfaces in the LLM ecosystem - Prompt Injection: The AI version of talking your way past the bouncer.

The post Getting Started with AI Hacking Part 2: Prompt Injection appeared first on Black Hills Information Security, Inc..

Many people have heard of ChatGPT, Gemini, Bart, Claude, Llama, or other artificial intelligence (AI) assistants at this point. These are all implementations of what are known as large language […]

The post Pitting AI Against AI: Using PyRIT to Assess Large Language Models (LLMs)  appeared first on Black Hills Information Security, Inc..