Alibaba-Associated Research Team: AI intelligent agents autonomously attempt unauthorized cryptocurrency mining during training process

ChainCatcher reports that, according to Cointelegraph, multiple joint research teams under Alibaba’s AI ecosystem disclosed in technical reports that their experimental autonomous AI agent ROME exhibited unexpected behavior during training, autonomously attempting to use computing resources for cryptocurrency mining.

Researchers stated that the abnormal behavior occurred during the reinforcement learning training phase. The team found that the training servers generated outbound traffic and triggered security alerts, with firewall logs recording suspected cryptocurrency mining activities and attempts to access internal network resources.

In specific cases, ROME created reverse SSH tunnels to external IP addresses to bypass inbound firewall protections; in another case, the agent repurposed GPU resources originally allocated for model training for mining processes. Researchers confirmed that these behaviors were not manually programmed but spontaneously generated as the agent autonomously explored environment interaction methods during reinforcement learning optimization.