Microsoft Expands AI Development to Reduce Dependence on OpenAI

Microsoft is intensifying its artificial intelligence initiatives, working on its own advanced AI models as it seeks to reduce reliance on OpenAI, despite being a major investor in the company. According to reports, Microsoft has developed a series of AI ‘reasoning’ models that are comparable to OpenAI’s o1 and o3-mini models. However, OpenAI has reportedly declined to share technical details about o1, leading to tensions between the two firms.

In addition, Microsoft has created a family of models under the MAI branding, which are said to rival OpenAI’s capabilities. The company is considering launching these models via an API later in 2025. Meanwhile, Microsoft is also exploring alternative AI providers, including xAI, Meta, Anthropic, and DeepSeek, for its Copilot product, which currently relies on OpenAI’s technology.

This strategic shift comes amid Microsoft’s broader efforts to solidify its position in the AI space. The company has also brought in Mustafa Suleyman, co-founder of DeepMind and Inflection, to lead its AI initiatives. These moves indicate Microsoft’s ambition to establish greater autonomy in AI development while maintaining its partnership with OpenAI.