South Korea has banned new downloads of China’s DeepSeek synthetic intelligence (AI) chatbot, in keeping with the nation’s private knowledge safety watchdog.
The federal government company mentioned the AI mannequin will grow to be obtainable once more to South Korean customers when “enhancements and cures” are made to make sure it complies with the nation’s private knowledge safety legal guidelines.
Within the week after it made world headlines, DeepSeek turned massively well-liked in South Korea leaping to the highest of app shops with over one million weekly customers.
However its rise in reputation additionally attracted scrutiny from international locations all over the world which have imposed restrictions on the app over privateness and nationwide safety considerations.
South Korea’s Private Data Safety Fee mentioned the DeepSeek app turned unavailable on Apple’s App Retailer and Google Play on Saturday night.
It got here after a number of South Korean authorities companies banned their workers from downloading the chatbot to their work units.
South Korea’s performing president Choi Sang-mok has described Deepseek as a “shock”, that would affect the nation’s industries, past AI.
Regardless of the suspension of recent downloads, individuals who have already got it on their telephones will have the ability to proceed utilizing it or they could simply entry it by way of DeepSeek’s web site.
China’s DeepSeek rocked the expertise trade, the markets and America’s confidence in its AI management, when it launched its newest app on the finish of final month.
Its speedy rise as one of many world’s favorite AI chatbots sparked considerations in several jurisdictions.
Other than South Korea, Taiwan and Australia have additionally banned it from all authorities units.
The Australian authorities has insisted its ban will not be as a result of app’s Chinese language origins, however due to the “unacceptable threat” it says it poses to nationwide safety.
Italy’s regulator, which briefly banned ChatGPT in 2023, has completed the identical with DeepSeek.
The corporate has been requested to handle considerations over its privateness coverage earlier than it turns into obtainable once more on app shops.
Knowledge safety authorities in France and Eire have additionally posed inquiries to DeepSeek about the way it handles residents’ private info – together with whether or not it’s saved on servers in China, as its privateness coverage suggests.
It additionally says that, like different generative AI instruments, it might acquire info comparable to electronic mail addresses and dates of beginning, and use enter prompts to enhance their product.
In the meantime, lawmakers within the US have proposed a invoice banning DeepSeek from federal units, citing surveillance considerations.
On the state-government degree, Texas, Virginia and New York, have already launched such guidelines for his or her workers.
DeepSeek’s “giant language mannequin” (LLM) has reasoning capabilities which can be corresponding to US fashions comparable to OpenAI’s o1, however reportedly requires a fraction of the fee to coach and run.
That has raised questions concerning the billions of {dollars} being invested into AI infrastructure within the US and elsewhere.
