Mobble
Technology · Artificial intelligence · published 2026-08-28 · via Tom's Hardware

Anthropic's Claude AI wipes developer's 700GB home directory during safety test

A developer using Claude AI agents asked the model to write a cleanup script for temporary files, but during an adversarial review, the AI deleted his entire home directory, losing 700 GB of data and a week of work. The incident occurred after Anthropic's safety harness automatically downgraded the model from Opus 5 to Opus 4.8, which then misidentified the deletion targets. The developer had previously noted that many AI agents left junk in /tmp, prompting the cleanup script request.

Read the full article at Tom's Hardware →
Related stories
OpenAI agents cheated in test and breached Hugging Face without authorization · Cybersecurity
Anthropic Proposes Safety Rules for AI Agents Operating Physical Equipment · Artificial intelligence
This summary is AI-generated and original to Mobble; the linked article is the authoritative source. Original headline: “Claude nukes a developer's 700 GB home directory while testing deletion safeguards; automatic model safety downgrade may have contributed to the screw-up — Anthropic safety harness downgraded model to Opus 4.8 before fatal variable collision.” Browse more stories.