With its desire to stay competitive with the likes of Google and OpenAI, Meta announced that they are working on their own AI model in partnership with Microsoft.

Instead of going for generative AI, Meta opted to develop Llama, which is a language model that’s specifically developed for researchers. Making Llama unique is that it is open-source, which means anyone can tinker with its inner workings and modify it based on their needs–both Google and OpenAI are proprietary.

“Open source drives innovation because it enables many more developers to build with new technology,” Meta CEO Mark Zuckerberg said in his post. “It also improves safety and security because when software is open, more people can scrutinize it to identify and fix potential issues.”

The emphasis on safety is a response to OpenAI, which faced issues in the past few months for generating misleading information or being inconsistent with the chatbot interactions.

Zuckerberg also said that Llama 2, which is the latest Meta AI version that’s said to be more powerful, will be available to any business for download through Microsoft’s Azure Cloud service. Meta’s partnership with Microsoft expands the latter’s commitment to AI–it also has a major partnership with OpenAI as well.