The US Chatgpt developer Openai and his partner Microsoft According to a report, the suspicion of a data clause Chinese competition after. Security experts from Microsoft had observed the drainage of large amounts of data via an interface of the Openai software last autumn, the news agency wrote Bloomberg on Tuesday (local time). Those responsible for this would presumably be with the Chinese start-up DeepSeek in connection.
Its cost -effective and yet technologically high -quality artificial intelligence (AI) had excited the industry at the beginning of the week.
Information from the AI models distilled
David Sacks, who is the AI policy of the US President Donald Trump Coordinated, the theft described intellectual property in a TV interview as possible. “There is valid evidence that Deepseek has distilled information from the AI models from Openai.” On request, Openaai was similar and emphasized that Chinese and other companies constantly tried to copy the AI programs of the leading developers from the USA. Openaai did not name the name Deepseek.
In the so-called distillation, existing AI models are used to train new or further developed models. So far, providers have already improved their own models. What is new is that Deepseek is said to have used information from foreign models to train your own AI.
Openai wants to protect intellectual property with the US government
“We take countermeasures to protect our intellectual property,” said the Chatgpt developer. “It is crucial for the future that we work closely with the US government in order to protect the most powerful models from the efforts of opponents and competitors to take over US technologies.” Microsoft did not want to comment on the possible economic espionage. Deepseek was initially not available for a comment.