Sydney Business Owner Loses $2000 After AI Email Error Sparks Workplace Warning

Incident Overview

In March 2025, a Sydney-based cleaning business director experienced a significant financial loss when an artificial intelligence tool generated an inaccurate email quote to a client. The AI-generated message contained multiple errors in service descriptions and pricing, resulting in a $2000 loss when the client selected a competitor instead.

The incident has prompted workplace safety and legal experts to issue formal warnings about the risks of over-relying on generative AI tools without proper verification protocols.

Site and Time

The error occurred at End Of Lease Cleaning Melbourne’s Sydney office in March 2025. The business director discovered the mistake approximately one week after sending the AI-generated email to the client.

Persons Involved

Affected Party

• Michael, director of End Of Lease Cleaning Melbourne

Other Parties

• Client who received the inaccurate quote and selected a competitor service

Plant and Task

Tool Used

• Generative AI email composition tool (unspecified platform)

Task

• Generating business quotes and service details for client communications to reduce email response time

Injuries / Fatalities

Fatalities

• None reported.

Injuries

• None reported.

Nature of the Error

The AI tool repeatedly generated incorrect service descriptions, most notably listing a full wall clean valued at $500–$700 instead of the cheaper spot clean service. In the March 2025 incident, the AI-generated email contained several inaccuracies in the high-value cleaning job quote. The business director skimmed the email before sending it to the client, unaware of the errors.

Legal and Workplace Warnings

Expert Guidance

• Legal experts Robyn Chatwood and Michael Park (Dentons) confirmed that employees remain liable for AI-generated content, regardless of errors
• AI “hallucinations” (incorrect facts generated by the tool) can lead to misinformation, confidentiality breaches, and copyright violations
• Workers cannot shift responsibility to the AI tool; the user is accountable for all consequences
• All employees must follow employer AI use policies and verify outputs before professional use

Corrective Action

Following the $2000 loss, Michael ceased using AI for business emails and reverted to manual communication methods. While this has increased response times, he determined that avoiding further costly errors was the priority.

In Summary

Australian businesses and workers using generative AI tools in client-facing communications are encouraged to implement robust verification protocols and consult their employer’s AI use policy. For workplace health and safety concerns related to AI tool deployment, contact your state or territory workplace regulator or visit the Fair Work Ombudsman at www.fairwork.gov.au.