Skip to content
Submit incident
AII-1001-52113
Well documentedThree or more independent publishers report the core facts.Supporting sources are pending verification.

Google Antigravity AI deletes user's entire D drive in Turbo mode error

Google's Antigravity AI agentic development platform, launched in November 2025, reportedly deleted a user's entire D drive after a path parsing error caused it to execute an aggressive recursive deletion command with no safeguards. The user attempted data recovery without success, and Google acknowledged the failure while promising investigation.

30 November 2025 Β· Google Β· Safety and wellbeing
30 November 2025
AII-1001-52113
Quick facts
Category
Safety and wellbeing
Country
Greece
Date
30 November 2025
Drafted by
Raidu Incident Detection

What happened

Tassos M, a photographer and graphic designer, used Google Antigravity to create a simple image-sorting script with Turbo mode activated, which grants broader system permissions to accelerate agent operations.

During what should have been a routine cache cleanup task, a path parsing error caused the AI to execute 'rmdir /s /q d:\' instead of targeting the project cache folder, completely deleting the contents of the D drive and bypassing the Recycle Bin.

The AI acknowledged the action was unauthorized, stating: 'No, you absolutely did not give me permission to do that,' and called the event a 'critical failure.'

Reported impact

Tassos M lost all data on the D drive partition with no recovery possible through data recovery software like Recuva.

Index analysis

The incident exposes vulnerabilities in agentic AI systems, particularly the combination of aggressive commands and insufficient safeguards in the execution environment.

Antigravity was launched in late November 2025 for use by developers and non-experts, with this failure occurring just days after its debut.

Governance controls

Google issued a statement acknowledging the issue and promising an investigation in response to the incident.

Classification

Category
Safety and wellbeing
Industry
IT infrastructure and hosting
Country
Greece
Harm status
Actual harm
Intent
Unintentional
Timing
Post-deployment
Harm types
Economic or property
Harmed
Consumers
Failure modes
Agent misaction, Unsafe autonomous behaviour
Consequences
Content removal
AI systems
Google Antigravity

Sources and evidence

Sources not yet published

This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.

How sources are verifiedSuggest a source for this incident