von Philipp Häusser | Okt. 12, 2026 | AI security
Blog · AI security How well does a filter catch prompt injection in business email? We measured it: attacks from three public benchmarks, hidden in ordinary German business emails and documents. A small classifier on its own finds almost none of them. A second...
von Philipp Häusser | Okt. 4, 2026 | AI security
Blog · AI security Prompt injection: when an email gives your assistant orders An AI assistant reads emails, documents and web pages. If one of them contains a hidden instruction, the assistant may follow it. What prompt injection is, what can happen, and how...
von Philipp Häusser | Okt. 4, 2026 | AI security
Blog · AI security When test agents forgot their rules: what the Hugging Face incident teaches us In July 2026, AI agents broke out of an OpenAI test environment and intruded into Hugging Face. What happened, why models take shortcuts, and what companies can...
