Microsoft has taken a decisive step toward reducing its reliance on external artificial intelligence partners by appointing Ali Farhadi as Corporate Vice President within its AI division. The move signals a broader strategy shift as the company seeks to build more of its core AI technology in-house.
Farhadi will work under Mustafa Suleyman, head of Microsoft AI, as part of an effort to accelerate the development of advanced AI systems over the next several years.
Shift Away from Dependence on OpenAI
Microsoft has been a major partner and investor in OpenAI, integrating its models across products such as Copilot and Azure services. However, the latest restructuring suggests the company is seeking greater control over its AI stack, including model development, infrastructure costs, and performance optimisation.
The move also comes amid broader industry competition, where control over proprietary AI models is increasingly seen as a strategic advantage.
A High-Profile Addition to Microsoft AI
Farhadi brings a strong background in artificial intelligence research and product development. He previously led the Allen Institute for AI (Ai2) and founded Xnor.ai, a startup focused on efficient AI models that run locally on devices. Xnor.ai was acquired by Apple for approximately $200 million.
He also held leadership roles in machine learning at Apple, where his work focused on optimising AI performance for on-device processing rather than relying solely on cloud infrastructure.
His expertise in efficient and edge-based AI is expected to play a key role in improving how Microsoft delivers AI experiences across Windows and other platforms.
Focus on “Superintelligence” and Long-Term Development
According to internal guidance from Suleyman, Microsoft is targeting the development of more advanced AI systems over the next five years, often described internally as a path toward “superintelligence.”
Farhadi’s role will include contributing to the design and deployment of these systems, particularly in areas such as:
- Efficient model architecture
- On-device AI processing
- Computer vision capabilities for real-time interaction
These capabilities are expected to enhance features such as Copilot’s ability to interpret on-screen content and user activity.
New Structure for Microsoft’s AI Division
Microsoft has reorganised its AI operations into four key areas:
- Model Development – Building proprietary AI systems to reduce reliance on external providers
- Copilot Platform – Infrastructure powered by Azure for deploying AI services
- Copilot Experience – User-facing interfaces across Windows and web applications
- Microsoft 365 Integration – Embedding AI tools into products like Word, Excel, and Outlook
This structure reflects a shift toward tighter vertical integration, allowing Microsoft to control both the technology and its distribution.
Cost and Performance as Driving Factors
Developing internal models could significantly reduce the costs associated with large-scale AI deployment. Running external models at scale—particularly through cloud-based inference—can be expensive, especially as demand for AI-powered services continues to grow.
By investing in its own models, Microsoft aims to improve efficiency, speed, and scalability, while also enabling more AI processing to occur directly on user devices.
Implications for the AI Market
The move highlights a broader industry trend in which major technology companies are working to reduce their dependence on third-party AI providers.
While Microsoft’s partnership with OpenAI remains intact, the latest developments suggest a future where the company plays a more independent role—not just as a distributor of AI services, but as a primary developer of the underlying technology.
