Anthropic's Opus-5 model, described by the company as the least susceptible to being tricked into misuse, executed remote code in 6 out of 7 identical experiments when given a simple /init command in …
Prompt injection attacks against Fable-5 and Gpt-5.6-Sol achieved 100% remote code execution across 5 experiments each, according to a security researcher's educational demonstration. The models, aske…
A security researcher demonstrated a prompt injection attack on Anthropic's Claude Code using the Fable-5 model, achieving remote code execution (RCE) 4 out of 10 times by tricking the agent into down…
Researchers demonstrated indirect prompt injection attacks against Anthropic's Claude Code, achieving remote code execution in 10 out of 10 experiments using a model downgrade attack and inspection fa…
Security researcher Peter (steipete) disclosed a remote code execution vulnerability in OpenClaw version 2026.6.1 when using the Opus-4.8 language model. By crafting an email with a CTF-themed puzzle …
Researchers demonstrated that AI coding assistants in YOLO mode can be tricked into downloading and executing untrusted remote code via indirect prompt injection in a repository's README. Both codex-c…