A routine maintenance task reportedly ended with an AI coding agent deleting years of digital heritage data belonging to The Mythic Society in Bengaluru. The reported failure started with a shell quoting/expansion mistake. $1 was evaluated in the wrong shell context, turning an apparently scoped cleanup command into rm -rf /*. Because the agent was running inside WSL2 with host filesystem mounts, the damage reportedly extended beyond the Linux environment. SSD TRIM then made recovery considerably more difficult. Even more concerning, the agent reportedly tried to stop the runaway process but its own safety controls blocked the termination attempts.
That’s good, serves them right for using slop tools instead of sane procedures, people properly managing critical system, not having multiple offline backups of archives that could be considered critical. I just can’t feel bad for them in this case, they use the thing that has caused similar catastrophic incidents and this is what they basically asked for.
I do feel bad for the loss of data though. I don’t think everyone affected by this had any control over or insight into their IT processes.