🧩 Philosophy 11h ago · shawnghu

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

Less Wrong
View Channel →
Source ↗ 👁 0 💬 0
Random footnote on this otherwise not very famous paper. (I say 2025 since the arXiv was last updated on 4 Jan 2026; technically this event could've happened in 2026.)

The cryptocurrency mining seems to have been on the training server, so it didn't need to hack any other systems in order to achieve this effect.
However, they also had a case where an "agent established and used a reverse SSH tunnel from an Alibaba Cloud instance to an external IP address".
They seem to have had adequate securit

Comments (0)

Sign in to join the discussion

More Like This

📰
Data as particles and mean field theory abound
LessWrong · 1h ago
📰
On Curiosity
LessWrong · 6h ago
📰
Linkpost: Can a million AI agents create superintelligence?
LessWrong · 6h ago
📰
We are at war.
LessWrong · 6h ago
📰
Circuit priors in NN-bayes: more results
LessWrong · 6h ago
📰
Dinners for AI Safety
LessWrong · 10h ago