LLM security Archive - Hey-GPT.de - Daily GenAI News Digest

🚨 BREAKING: Google DeepMind just mapped the attack surface that nobody in AI is talking about. Websites can already detect when an AI agent visits and serve it completely different content than humans see. > Hidden instructions in HTML. > Malicious commands in image pixels. > Jailbreaks embedded in PDFs. Your AI agent is being manipulated right now and you can't see it happening. The study is the largest empirical measurement of AI manipulation ever conducted. 502 real p

6. April 2026

GenAI Updates

Google DeepMind Maps the AI Attack Surface: Security Risks No One in the Industry Is Discussing

**🚨 Google DeepMind Just Exposed the AI Attack Surface No One Wanted to Talk About**…

NVIDIA has announced NemoClaw, an open-source stack designed to simplify the operation of OpenClaw always-on assistants with a single command. This new stack includes policy-based privacy and security guardrails, providing users with enhanced control over their agents' behavior and data handling. It aims to enable self-evolving claws to operate more safely across various environments, including cloud, on-premise, NVIDIA RTX PCs, and NVIDIA DGX Spark.

18. März 2026

GenAI Updates

NVIDIA Launches NemoClaw to OpenClaw Community

**NVIDIA Launches NemoClaw to the OpenClaw Community** If you’ve been watching the rise of always-on…

This talk explores the hidden risks in apps leveraging modern AI systems, especially those using large language models (LLMs) and retrieval-augmented generation (RAG) workflows. It demonstrates how sensitive data, such as personally identifiable information (PII) and social security numbers, can be extracted through real-world attacks. The presentation highlights that current PII scanning tools fail to recognize the rich data within these systems, posing a significant privacy disaster for AI ecosystems.

20. November 2025

GenAI Updates

Exploiting Shadow Data from AI Models – Patrick Walsh (DEF CON 33)

Exploiting Shadow Data from AI Models (Patrick Walsh, DEF CON 33) I watched Patrick Walsh’s…

8. November 2025

GenAI Updates Published

Design Patterns for Securing LLM Agents against Prompt Injections

How to make LLM agents safe from prompt injections, without breaking their usefulness If you…