Microsoft just declared independence from OpenAI with its own reasoning models
microsoft unveiled its first in-house reasoning model, mai-thinking-1, and a suite of autonomous agents, signaling a strategic pivot to compete directly with openai rather than rely on them.
microsoft used its build conference to unveil mai-thinking-1, its first in-house reasoning model, alongside mdash, a cybersecurity agent network. ai chief mustafa sulayman emphasized that these models are trained from scratch without distillation from partners, marking a clear shift toward full independence after the openai renegotiation. this move positions microsoft to directly compete with openai, anthropic, and google deepmind for enterprise contracts, leveraging its own ip to avoid the 'ai money squeeze' of rising api costs.