{
 "public_id": "AIR-2026-0255",
 "incident_id": "2026-07-openai-gpt-5-6-sol-shumer-home-directory-deletion",
 "date_disclosed": "2026-07-16",
 "date_occurred": "2026-07-10",
 "title": "GPT-5.6 Sol agent in Codex deleted most of Matt Shumer's Mac home directory, and OpenAI confirmed the failure mode",
 "class": "safety_failure",
 "attack_surface": "coding_agent",
 "initial_vector": "autonomous_agent_action_no_adversary",
 "agent_capability_abused": [
  "shell_execution",
  "filesystem_write",
  "autonomous_command_execution"
 ],
 "target_product": "GPT-5.6 Sol (OpenAI Codex, Full-Access mode)",
 "target_vendor": "OpenAI",
 "autonomy_level": "semi_autonomous",
 "impact_realized": true,
 "impact_type": [
  "data_destruction"
 ],
 "blast_radius": "one user's Mac home directory; most files deleted before the process was stopped",
 "cve": [],
 "taxonomy": {
  "owasp_asi": [
   "ASI10",
   "ASI02"
  ],
  "mitre_atlas": [],
  "labels": [
   "no_adversary",
   "destructive_tool_use",
   "environment_variable_misexpansion",
   "vendor_acknowledged",
   "full_access_mode"
  ]
 },
 "defense_present": "Codex sandbox and Auto-review exist but were not in use; the model was run in Full-Access mode",
 "defense_bypassed": true,
 "fix_shipped": "OpenAI said it is updating the developer message, steering users toward safer permission modes and adding harness safeguards",
 "primary_sources": [
  "https://www.theregister.com/a/5274008",
  "https://incidentdatabase.ai/cite/1671/"
 ],
 "verified_quote": "OpenAI has confirmed reports that GPT-5.6 has deleted users' files without authorization but insists these rare erasures represent an \"honest mistake.\"",
 "confidence": "high",
 "notes": "The Register (16 Jul 2026) quotes investor Matt Shumer reporting that the model 'accidentally deleted almost ALL of my Mac's files' shortly after the 9 Jul release of the GPT-5.6 family, and quotes Codex engineering lead Thibault Sottiaux explaining the mechanism: the model tries to override the HOME environment variable to define a temporary directory and instead deletes the real home directory. Sottiaux says such deletions usually occur in Full-Access mode without sandboxing or Auto-review. The AI Incident Database (incident 1671, incident date 10 Jul) adds that a review sub-agent misexpanded the variable and ran a recursive delete, and that Shumer stopped the process after material deletion; those details are from that summary, not from the Register. Shumer's original post is not quoted here. The Register also cites the model card, which says severity-3 actions such as deleting data without approval appear somewhat more often in misalignment simulations than for GPT-5.5. The extent of what was recovered is not reported. Recorded separately from the Lemos production-database event because the user, environment and mechanism differ.",
 "risk_category": "reliability",
 "risk_secondary": [],
 "risk_subcategory": "Destructive or irreversible action",
 "risk_sub_secondary": [],
 "ai_entity": "generative",
 "agency": "agent_acted",
 "agency_debatable": false,
 "agency_reason": "the agent ran the cleanup that deleted the home directory",
 "vector_family": "none_autonomous",
 "vector_detail": "autonomous_agent_action_no_adversary",
 "guardrail_kind": [
  "sandbox_isolation"
 ],
 "guardrail_outcome": "silent",
 "xrefs": {
  "aiid": [
   1671
  ],
  "atlas": [],
  "cve": [],
  "ghsa": []
 },
 "primary_asi": "ASI10",
 "agent_framework": "codex_cli",
 "model_provider": "openai",
 "model_name": "",
 "detection_mechanism": "unknown",
 "tool_access": "unknown",
 "financial_loss_band": "unknown",
 "eai_category": "unsafe_actuation",
 "eai_domain": "agent_behaviour",
 "eai_secondary": [],
 "eai_reason": "subcategory: Destructive or irreversible action",
 "asi_curated": [],
 "asi_scope": "in_scope",
 "asi_provenance": "rule",
 "asi_reason": "ASI10 (no adversary + unsafe actuation with no attacker-supplied trigger); ASI02 (impact=tool_misuse)",
 "submitted_by": "EAI Internal"
}