OpenAI tightens the screws on security to keep away prying eyes

Openai said he had overhauled security operations to protect corporate spying. According to the financial timeThe company accelerated the existing security squeezing after the beginning of China.

The safety of beef provision includes the policy of a “information tent”, which is limited to the security of security, sensitive algorithms and staff. For example, during the development of the Openai O1 model, the only approved team members read in the project can discuss it in the shared office.

And there are more things. Now Openai isolates special technology in offline computer systems, and the biometric access controls in the office, which requires an open confirmation to foreign relations, and this is the company’s information centers, and added physical security in the information centers and expanding the staff of physical security.

If we take into account the ongoing, Openai said that Openai reflects more concerns about foreigners trying to steal the intellectual property Breakyer Wars American AI companies and increasingly often occur leakage CEO Sam Altman’s comments, Openai can also try to solve internal security issues.

We reached Openai for comment.