Major Sears Samantha Data Breach Exposes AI Privacy and Surveillance Risks

U.S. retailer Sears deployed an AI customer-service system called Samantha, which uses voice activity detection, or VAD, to determine call status and process customer conversations. Customer-service data can contain names, contact details and purchasing information. If companies fail to properly restrict recording and storage permissions, AI tools can shift from service channels into sources of privacy and surveillance risk.

The latest disclosures show that the Samantha security incident exposed more than 3.7 million chat records and 1.4 million audio files. Investigators also found that the VAD system may have continued recording after calls ended, broadening the scope of unintended data collection. Reports gave no exact date for either the incident or its public disclosure and cited no financial amount, but the scale of the breach has raised concerns that companies are rushing to adopt AI without keeping cybersecurity governance in step.