AI Security

Security risks in artificial intelligence and machine learning systems, including LLM vulnerabilities and AI-generated code safety.

We Want AI. We Just Can't Say Why

We Want AI. We Just Can't Say Why

· 13 min read

In Denmark, AI data centres have queued for more power than the entire national grid can supply, with ten-year waits and almost no jobs. The power, the money, and a question we keep dodging: do we actually need this?
LLMs Can Write Code, but Cannot Read Your Mind

LLMs Can Write Code, but Cannot Read Your Mind

· 9 min read

LLMs generate valid code quickly, but without proper context they cannot distinguish between secure and insecure patterns. Capability without context is a threat to AI code quality, just like any external contribution requiring review.