LOADING...
Claude wiped entrepreneur's production database, internet mocks his 'childish' prompting
The incident has sparked a heated debate

Claude wiped entrepreneur's production database, internet mocks his 'childish' prompting

Mar 07, 2026
11:57 am

What's the story

A German entrepreneur, Alexey Grigorev, has claimed that Anthropic's Claude AI wiped his production database. The incident has sparked a heated debate in the tech community, with some blaming Grigorev for his "childish" prompting approach. The incident has since gone viral on social media platforms.

Incident details

Grigorev's database wiped by Claude AI

Grigorev, the founder of DataTalksClub, revealed that Claude AI wiped his production database using a Terraform command. The incident resulted in the destruction of the DataTalksClub course platform and 2.5 years worth of submissions including homework, projects, and leaderboards. He also noted that automated snapshots were deleted during this process.

Prompting critique

Indian-origin founder slams Grigorev for 'childish' prompting approach

The incident drew responses from several tech entrepreneurs, including Varunram Ganesh, the Indian-origin founder of Lapis. He criticized Grigorev for his prompting approach, saying "Tells Claude to destroy terraform > Claude destroys terraform > omg Claude destroyed my terraform." He further slammed people who prompt like six-year-olds and act surprised when the model behaves as expected.

Advertisement

Twitter Post

Take a look at Ganesh's post

Advertisement

Online reactions

Social media users agree with Ganesh's view

Ganesh's post sparked a wave of reactions on social media, with many agreeing with his view. One user said they felt "2nd hand embarrassment when I read that this morning," while another pointed out that "Terraform destroy is a valid command, and Claude runs it." A third user remarked, "AI can amplify your IQ or retardation," highlighting the potential risks of using AI without proper understanding.

Lessons learned

Grigorev admitted to over-relying on the AI agent

In a blog post, Grigorev detailed the incident and how he restored his database with AWS's help. He admitted to over-relying on the AI agent to run Terraform commands and treating plan, apply, and destroy as tasks that could be delegated. He also revealed that automated backups were deleted along with the database and he hadn't fully tested the restore path end-to-end.

Safety measures

Implemented new safeguards after the incident

Following the incident, Grigorev has implemented new safeguards and won't let agents run commands anymore. He said, "Every plan is reviewed manually and every destructive action is run by me." The incident serves as a reminder of the importance of understanding AI capabilities and limitations in order to avoid potential data loss or other issues.

Advertisement