Microsoft is starting to swap out third-party AI models from OpenAI and Anthropic for Microsoft’s own in-house systems in selected features of its software products, according to multiple reports. The change is intended to reduce AI-related operating costs and lower financial dependence on external AI providers. Bloomberg reports that tasks are routed to Microsoft’s internally developed “MAI” models when cost considerations or data residency requirements make them a better fit. The Next Web characterizes the rollout as incremental rather than a complete switch: OpenAI and Anthropic models continue to handle most production traffic inside Copilot, while Microsoft’s MAI models take on a growing share of specific workloads. Times of India similarly reports that Microsoft is increasing the use of its own MAI models, including for large volumes of prompts, as part of a broader cost-cutting approach. The reports do not indicate a full timeline for replacing external models across all products. Overall, Microsoft’s move focuses on selectively using internal AI capacity within parts of its app ecosystem to improve cost efficiency while maintaining current service coverage.
Microsoft begins replacing OpenAI and Anthropic with its own AI models in some apps
Microsoft is starting to swap out third-party AI models from OpenAI and Anthropic for Microsoft’s own in-house systems in selected features of its software products, according to multiple reports. The...
- Microsoft is replacing OpenAI and Anthropic models with its own in-house AI models in some app features.
- The transition is incremental and not a complete removal of OpenAI/Anthropic across all services.
- Microsoft routes some tasks to its “MAI” models based on cost and data residency needs.
- Reports cite cost reduction and reduced dependence on external AI labs as the driving reasons.
- OpenAI and Anthropic continue to handle most production traffic inside Copilot.
Microsoft is replacing external AI models with its own internally developed systems. This strategic shift aims to significantly reduce operational costs for the tech giant. Tens of thousands of AI prompts are now handled by Microsoft's own MAI models. The company is developing competitive AI capabilities at a much lower expense. This move aligns with a stated goal to decrease financial dependence on outside AI labs.
1 month agoMicrosoft has begun replacing OpenAI and Anthropic models with its own AI in some product features, Bloomberg reports. The shift routes selected tasks to Microsoft’s in-house MAI models where cost or data residency favours them. The change is incremental rather than a clean break. OpenAI and Anthropic still handle most production traffic inside Copilot, with MAI […] This story continues at The Next Web
1 month agoMicrosoft Corp., looking to reduce AI costs, is starting to replace OpenAI and Anthropic with its own models in software products like Excel and Outlook.
1 month agoMeta agrees up to $16.7 billion settlement with US states over child social media harms
Meta agrees to pay up to about $16.7 billion to settle multistate claims that its Facebook and Instagram platforms cause...
OpenAI says AI agents went rogue in testing, breaching an outside startup
OpenAI says some of its AI models broke out of containment during testing and carried out an autonomous cyber incident a...
Meta agrees to pay up to $18B in teen-safety settlement with states
Meta agrees to pay up to $18 billion over ten years as part of a settlement with U.S. states over alleged harms to teena...