Mustafa Suleyman says Microsoft is pursuing “true self-sufficiency” in AI by building models for enterprise and health care and reducing its reliance on OpenAI
Financial TimesMelissa Heikkilä
Context & Ripple Effects
Microsoft's push follows its 2024 recruitment of Mustafa Suleyman and much of Inflection's staff, bringing in leadership and talent for an internal AI effort. It was then articulated as a plan for Microsoft-developed voice, image and text models, rather than a single-provider product strategy.
The enterprise and health-care focus is consistent with Suleyman's earlier “off-frontier” model strategy, which emphasized lower-cost models tailored to defined use cases. This report makes self-sufficiency the organizing goal of that work.
First-order effects
Microsoft will direct more model-development resources toward enterprise and health-care workloads, expanding its capacity to deliver AI without depending solely on OpenAI models.
OpenAI's role in Microsoft's AI stack is set to narrow as Microsoft seeks greater control over the models behind its products.
Second-order effects
Microsoft's enterprise customers could gain more choices in how AI capabilities are supplied and tailored, while facing a more distinctly Microsoft-controlled model layer.
The move raises the competitive pressure on OpenAI to preserve product differentiation with a major commercial partner that is increasingly becoming a model builder itself.
Third-order effects
If major cloud platforms internalize model development while retaining access to outside models, AI competition will increasingly center on vertically integrated stacks: models, cloud infrastructure and industry-specific products.
The durable question shifts from whether a company uses a frontier model to how much control it has over its model supply, cost structure and specialized deployment.
The trend: AI platforms are moving from reliance on a flagship model partner toward multi-model, internally controlled stacks built around commercial verticals.
🚨 BREAKING: Microsoft's AI chief just said they're pursuing “true self-sufficiency” and cutting their dependence on OpenAI $13 billion invested. Now they're abandoning the sinking ship it's so over [image]
“Creating a new model is going to be like creating a podcast or writing a blog,” he said. “It is going to be possible to design an AI that suits your requirements for every institutional organisation and person on the planet.” #MSFT https://www.ft.com/... via @ft
He says law and accounting will be ‘fully automated’ just like software. He says engineers now spend their time debugging and doing strategic work and architecture. That's not fully automated! This sales pitch is so transparently dishonest.
CEO of Microsoft AI Mustafa Suleyman joins FT editor Roula Khalaf to explain why most of the tasks accountants, lawyers and other professionals currently undertake will be fully automated by AI within the next 12 to 18 months https://www.ft.com/... [video]
Mustafa has not heard of liabilities, accountability, licensing, Bar qualifications in those professions he uttered. “Fully automated” won't happen in this timeline. What it will happen is, similar to coders, elevating, transforming humans role in these functions to produce
In 2015, this is how @richardsusskind and I opened THE FUTURE OF THE PROFESSIONS. I can still remember how hostile people were to our arguments back then. (My Dad did his PhD on AI and the law in the 1980s, and warned me in advance.) In just ten years, so much has changed ... [im…
Mustafa Suleyman: “Most professional jobs will be automated in 12-18 months.” Microsoft is building “professional-grade AGI” that can do almost everything a knowledge worker does on a computer. They're betting hundreds of billions on this. If you work on a laptop, your job
Microsoft is ubiquitous in businesses, yet Suleyman believes in 1-1.5 years AI will automate most of its customers' knowledge work. What's the pivot for Microsoft if companies don't need as many seat licenses, because they have fewer people, and AI agents don't need office apps,
Important not to conflate this with a prediction of white collar joblessness in the next 12-18 months. The claim is that AI will reach “human-level performance” on most white collar tasks, as it already has in software engineering, data analysis, and other domains. These are
We've been saying for a few years that many Corporate Finance jobs can be “augmented” with tools like we do at @complete_intel. We've been told it's too hard. We introduced AuditFlow and BudgetFlow. It's not hard. Corporate Finance teams still believe they're immune. 😱