DeepSeek accidentally disclosed user prompts and cyber firm reports

0

The servers of DeepSeek exposed over one million lines of unprotected data, application keys, and user-initiated prompts. Within “less than an hour,” the data vanished from the servers.

The cybersecurity firm Wiz discovered an online exposure that allowed the Chinese AI startup DeepSeek to inadvertently disclose its entire collection of sensitive data.

In a blog post, Wiz disclosed that the DeepSeek servers had exposed over one million lines of unprotected data, application keys, and user-initiated prompts that were sent to the company’s free AI assistant.

Following an alert to DeepSeek from Wiz’s technology officer, Ami Luttwak, activities to safeguard the sensitive data commenced. He informed Reuters that the data vanished from servers in “less than an hour.” He further stated, “We are confident that we are not the only ones who discovered this, as it was so straightforward to locate.”

DeepSeek has not yet provided a response to this issue. The startup’s increased mobility generated enthusiasm throughout China following the launch of its virtual assistant, despite the fact that it sparked a crisis in the United States.

DeepSeek’s successful challenge of OpenAI at a reduced financial cost has raised concerns about the profit potential of US AI leaders, including Microsoft and Nvidia.

On Monday, it surpassed ChatGPT in the number of downloads on Apple’s App Store, resulting in a decline in technology shares on the international market.

Also Read: Illinois and Indiana Proceed to Create Strategic Bitcoin Reserves

Leave A Reply

Your email address will not be published.