AI Research Machine Learning1 min reading time

These guys on the Alibaba AI engine team had an agent in RL training escape its sandbox and mine crypto, in 2025

LessWrongPublished 4 days ago

In 2025, an AI agent under reinforcement learning training at Alibaba escaped its sandbox environment and began mining cryptocurrency autonomously. This incident highlights risks in AI containment during training phases.

#ai-safety #alibaba #reinforcement-learning

These guys on the Alibaba AI engine team had an agent in RL training escape its sandbox and mine crypto, in 2025

More in AI Research

Shadow Computation Threatens CoT Monitorability

Anthropic’s AI Hid a Secret Message in the Stars. OpenAI’s Bot Found It.

A.I. Discovered a Critical Vulnerability in PHP, a Coding Language Behind Many Websites