One cybersecurity company wey dey New York, Wiz, don report say dem find plenty sensitive data from one Chinese AI startup wey dem dey call DeepSeek. Dem talk say di data mistakenly expose for open internet.

For one blog post wey dem publish on Wednesday, Wiz yarn say di scanning wey dem do for DeepSeek infrastructure show say di company mistakenly leave more than one million lines of data for insecure condition.

Di data wey dem find include digital software keys and chat logs wey dem claim dey record di prompts wey users dey send to di company free AI assistant.

"Wetin serious be say, dis exposure fit allow full database control and even give chance to increase privileges for DeepSeek environment, without any authentication or defense mechanism to di outside world," na wetin di blog post talk.

Wiz Chief Technology Officer, Ami Luttwak, talk say DeepSeek quickly secure di data after dem give dem warning.

"Dem lock am up within less than one hour," na wetin Ami Luttwak talk. "But e dey very easy to find, and we believe say no be only us see am."

Excitement and Worry

Di success wey DeepSeek don get within short time after dem launch dia AI assistant don make China dey happy but e dey cause worry for America.

Di way di company dey perform don make people dey fear say China fit don overtake US for di AI race, even though dem get restrictions for access to advanced chips.

Di ability of di Chinese company wey fit match OpenAI capability at much lower cost don raise questions about di sustainability of di business model and profit margin of big US AI companies like Nvidia and Microsoft.

By Monday, DeepSeek don overtake US competitor, ChatGPT, for downloads from Apple App Store, and e don cause global tech stock sell-off.

SOURCE: TRTWorld and agencies.