OpenAI and Anthropic have agreed to provide the U.S. government’s National Institute of Standards and Technology (NIST) with early access to their upcoming AI models before they are publicly released.
Announced by NIST, this collaboration aims to help the newly formed US Artificial Intelligence Safety Institute assess potential safety risks associated with these AI models. This partnership enables NIST to evaluate and provide feedback on the models, contributing to safer AI development.
OpenAI has already begun complying with this agreement, having showcased a new AI model, “Strawberry,” to federal officials.
Anthropic, a company founded by former OpenAI employees, also emphasized the importance of this collaboration in enhancing the safety of their models. Both companies are committed to advancing responsible AI practices through these initiatives.
