PolicyTechCrunch

OpenAI institutes new safeguards after Hugging Face breach

#openai#security#hugging face#ai#monitoring

English

OpenAI has introduced new security policies aimed at improving the monitoring and alignment of AI models during development and post-training. These measures follow the Hugging Face breach and include enhanced network isolation and a monitoring system to detect unauthorized behavior.

中文

OpenAI推出了新的安全政策，旨在改善AI模型在开发和训练后阶段的监控和对齐。这些措施是在Hugging Face泄露事件后实施的，包括增强的网络隔离和监控系统，以检测未经授权的行为。