DeepSeek database exposes chat histories and personal data of 1 million users
A DeepSeek database was discovered exposed by Wiz Research, leaking chat histories, personal details, and operational information from approximately 1 million users of its AI systems. The exposure potentially allows attackers to impersonate users, access accounts, and commit identity theft or financial fraud.
What happened
Wiz Research discovered an exposed DeepSeek database containing chat histories and user interactions with DeepSeek AI systems, including R1 and V3 models, affecting approximately 1 million users.
The database also included details about DeepSeek's operations and information necessary to access its internal systems, allowing potential attackers to gain full control.
Reported impact
The exposure compromises user privacy and potentially allows malicious attackers to impersonate users or access their accounts, putting user data at immediate risk and opening possibilities for identity theft and financial fraud.
Index analysis
The breach appears to have occurred due to DeepSeek's rapid deployment of its generative AI products without taking adequate care about security.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.