Tigerfish data exposure left sensitive customer files accessible online without password protection. The San Francisco-based transcription company stored police interviews, body-camera recordings, audio and video files, and customer invoices in a publicly accessible Amazon S3 bucket.
The company serves law enforcement agencies, prosecutors and technology businesses. Tigerfish has since restricted access to the exposed storage.
Tigerfish data exposure reveals sensitive police records
Security researcher Anurag Sen discovered the exposed files in October 2026. The storage bucket allowed visitors to browse and access files through a publicly accessible web address.
Sen initially contacted Tigerfish but received no response. Journalists later alerted the company after examining the exposure.
Tigerfish President Adam Goldberg confirmed that the company had blocked public access to the affected bucket. He said Tigerfish had also checked its other S3 storage buckets and was reviewing its security practices.
According to the company, an email spam filter caught Sen’s original warning.
Customer notification and unauthorized access remain unclear
The exposed material included sensitive law enforcement records and transcription source files. Such records can contain information about investigations, witnesses and individuals who have not been charged with crimes.
Although the files were publicly accessible, no confirmed evidence shows that malicious actors downloaded them.
It also remains unclear whether Tigerfish notified affected customers or conducted a broader investigation into potential unauthorized access.
A separate cloud security incident previously affected Uber drivers. TechAmerica.ai reported on the exposure of driver identification records and subsequent customer notifications.
Cloud storage mistakes create recurring security risks
The Tigerfish incident illustrates how a cloud configuration error can expose sensitive information without a sophisticated cyberattack.
Similar access-control problems have affected software development platforms. In April 2026, Lovable acknowledged that a permissions regression allowed authenticated users to access source code and chat histories in certain public projects.
In its official incident response, Lovable said it corrected the problem and changed its project visibility controls. That incident was separate from Tigerfish’s storage exposure.
TechAmerica.ai also examined the Fireworks AI data breach involving Notion customer conversations, highlighting another risk associated with externally hosted information.
For organizations handling confidential records, storage access restrictions and regular security audits are essential. Clear procedures for receiving vulnerability reports can also help companies resolve exposures before sensitive data is compromised.










