News Summary

OpenAI has significantly strengthened its security measures in response to fears of corporate espionage, particularly after a Chinese startup, DeepSeek, allegedly copied its models. The company has implemented policies such as ‘information tenting’ to restrict staff access to sensitive information and proprietary technology. Additional security measures include offline isolation of technology, biometric access controls, and a ‘deny-by-default’ internet policy. These steps highlight OpenAI’s efforts to protect its intellectual property against foreign threats and internal leaks.