Sphere, openai and ai21 jointly publish the best practice guidelines for deployment models

Guidelines include ：

1.Prohibit misuse

2.Mitigate unintentional harm

3.Thoughtfully collaborate with stakeholders

In the guideline, it means , As LLM Provider , Publishing these principles is the first step towards safer co-operation guidance . Large scale language model development and deployment . The three institutions will continue to promote cooperation , And work with others , Reduce and prevent malicious use of language models .