June 16, 2024

Balkan Travellers

Comprehensive up-to-date news coverage, aggregated from sources all over the world

Samsung has reportedly leaked its private secrets through ChatGPT • The Register

Samsung has reportedly leaked its private secrets through ChatGPT • The Register

Less than three weeks after Samsung lifted a ban on employees using ChatGPT, the chaebol has reportedly leaked its secrets at least three times — including sensitive information on a semiconductor in development.

The ban was originally intended to protect company data, but was lifted on March 11 to boost productivity and keep employees engaged with the world’s latest cool tech gadgets.

According to local Korean media a report Those leaked secrets now include equipment metering and productivity data from the Device Solution, Agglomeration and Semiconductor business unit.

A staff member told the local outlet that they copied all the problematic source code of the semiconductor database download program, entered it into ChatGPT, and asked for a solution.

Another uploaded software code designed to identify faulty equipment, and a third uploaded records a meeting in an attempt to automatically generate minutes.

ChatGPT’s Instructions It states “Your conversations may be reviewed by our AI trainers to improve our systems.”

Therefore, Samsung’s secrets may be within the reach of OpenAI – an entity that has more than a passing interest in the technology and affairs of the Korean giant.

Once the incidents were discovered, Samsung reportedly implemented “emergency measures” that included limiting upload capacity to 1024 bytes per question.

“In the event of a similar incident even after taking emergency information protection measures, access to ChatGPT may be blocked on the corporate network,” Samsung bosses reportedly warned employees.

Chaebol had already warned its employees about leaking proprietary information when it lifted its blanket ban on ChatGPT. OpenAI similarly Warn Against downloading sensitive data.

See also  WhatsApp to allow users to edit messages within 15 minutes

Starting from March 1 this year, OpenAI policy dictate API users must opt-in to data sharing to train or improve their models, while non-API-dependent services—such as ChatGPT or DALL-E—require the user to opt-out to avoid their data being used.

“We remove any personally identifiable information from the data that we intend to use to improve form performance. We also only use small samples of data per customer for our efforts to improve form performance.” Claims Open AI.

Local media reports indicate that Samsung is now considering building its own AI service to prevent further accidents.

reg Samsung has asked to confirm the details of this story, but has not received a response at the time of writing.

according to Korea TimesAnd The incident rocked local tech companies, including chipmaker SK hynix and consumer device maker LG Display, which are now working on guidelines for the use of AI-powered chatbots.

SK hynix has reportedly banned the use of chatbots on its internal network and employees wishing to use the service must obtain security clearances. Obviously, LG Display decided to launch an educational campaign so that its employees understand how to protect company secrets.

The first step is supposed to be “don’t upload it to someone else’s website”. ®