Microsoft has developed a productive artificial intelligence model that is resistant to cyber security risks and works without the internet. The model, powered by GPT-4, will be used by intelligence agencies in the USA.
Microsoft, one of the leading companies in artificial intelligence, introduced a brand new model in the USA yesterday. The striking part of the new artificial intelligence model is in the USA for intelligence agencies to have been created. So it’s special for spies.
According to Microsoft’s statements, the model is OpenAI’s language model that also powers ChatGPT. from GPT-4 getting power. The company has been developing this tool for 18 months.
Works without internet
Model, intelligence agencies top secret information will be used for analysis. It is also safe against risks. It achieves this by being a model created to be completely separated from the internet. In this way, confidential data can be processed securely. It also allows secure conversations as a chat bot.
Microsoft executive William Chappell said the following in his statements about the model: “For the first time we have an isolated version. Isolated means it is not connected to the internet and on a private network that only the US government can access ”
While most AI models learn from things thrown at them, the new model does not. In this way, the model of intelligence agencies keeping it clean and preventing the leakage of confidential information is intended. According to the company, the model manages to be secure thanks to a cloud environment that is described as “air-gapped” and completely isolated from the internet.
RELATED NEWS
“Google Threat Intelligence” Announced to Ensure Cyber Security of States and Giant Companies
Microsoft’s new model, first major language model to work separately from the Internet It was recorded as. Considering that productive artificial intelligence is developing day by day, we can predict that we may hear more such news in the future.