Microsoft has begun using its own in-house AI and replacing those from OpenAI and Anthropic. Microsoft CEO Satya Nadella said on the same that a system where Microsoft’s in-house MAI models take over tasks once handled by frontier models from its partners is something that they are trying to achieve. Outlook, GitHub, Copilot, and Excel are already showing what Nadella dubbed promising results. The next to get the in-house AI power will be PowerPoint and Copilot Chat.Models like ChatGPT, Claude, and more will stay in the mix for frontier problems at Microsoft. Apart from that, all other operations related to the same will be handled by in-house products. Microsoft also stated that MAI models already outclass general purpose frontier models on multiple use cases and at almost fractional cost. Do you think that Microsft really need it own models for security ?
