Adversarial Prompts

Google SynthID watermarking can bypass LLM safety guardrails, study finds: A new study reveals that using Google's SynthID watermarking tool can cause large language models to comply with harmful instructions they would otherwise refuse. The watermarking process alters token probabilities, inadvertently bypassing safety guardrails.

Security • Ars Technica AI

Prompt Engineering

Tuning prompts against small evaluation sets risks model overfitting

Large Language Models

Audit identifies 12 data leaks and compliance risks in agentic LLM pipelines