Mustafa Suleyman plots AI ‘self-sufficiency’ as Microsoft loosens OpenAI ties
Big Tech group’s AI chief predicts white-collar work could be automated within 18 months
Microsoft is pursuing “true self-sufficiency” in AI by building its own powerful models and reducing its reliance on OpenAI, according to the company’s AI chief.
Mustafa Suleyman told the FT that the strategic shift follows a restructuring of its relationship with the ChatGPT maker in October. That has prompted the $3tn company to build its most advanced technology independently, rather than rely on an external partner.
“We have to develop our own foundation models, which are at the absolute frontier, with gigawatt-scale compute and some of the very best AI training teams in the world,” said the Google Deepmind co-founder who joined Microsoft in 2024.
The software giant is investing heavily in assembling and organising the vast datasets needed to train advanced systems. “That’s our true self-sufficiency mission,” he added.
Microsoft, one of OpenAI’s biggest and earliest investors, has relied on OpenAI’s models to power its own AI tools such as its Copilot software assistant.
Last year, it agreed to allow the start-up to complete a corporate restructuring, retaining a $135bn stake in the ChatGPT maker and keeping access to the group’s most advanced models until 2032.
But the deal also provides OpenAI greater freedom to seek new investors and infrastructure partners, potentially turning it into a more direct competitor.
Microsoft has spread its bets further, investing in other model makers such as Anthropic and Mistral. But the company has also accelerated the development of its own in-house models, which Suleyman said would launch “sometime this year”.
In particular, Suleyman said Microsoft’s goal is to capture more of the enterprise market by working on “professional grade AGI” [artificial general intelligence], meaning powerful AI tools that can complete daily tasks for knowledge workers.
“White-collar work, where you’re sitting down at a computer, either being a lawyer or an accountant or a project manager or a marketing person — most of those tasks will be fully automated by an AI within the next 12 to 18 months,” Suleyman said.
These AI agents will be able to co-ordinate better within the workflows of large institutions in the next two to three years, he added. The AI tools will also be able to learn and improve over time, taking more autonomous actions.
“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.”
However, Microsoft faces stiff competition in the enterprise market. Anthropic has gained a strong lead in AI-powered coding tools, while OpenAI and Google are also investing heavily in securing lucrative corporate AI deals.
Microsoft has forecast it will spend $140bn in capital expenditure in its fiscal year, which ends in June, as it increases spending on the infrastructure needed to build AI.
Investor fears that such spending is inflating an AI “bubble” have hammered Big Tech stocks, with Microsoft’s shares down more than 13 per cent over the past month.
“There’s no question these are unprecedented times, and I think markets are trying to wrap their head around how this plays out over the next five years,” said Suleyman. But he added: “We all have no doubt that these returns do compound to revenue and to bottom line.”
Suleyman said another focus for Microsoft was applying AI to healthcare in an effort to build “medical superintelligence”, where AI programs can help solve staffing crises and long waiting times for overstretched health systems. Last year, the company unveiled an AI diagnostic tool, which it claims can outperform doctors on some tasks.
He added Microsoft’s goal was to build “humanist superintelligence”, meaning AI technologies that remain under human control, addressing growing concerns that rival AI labs were rushing to build powerful technologies that resist the oversight of their creators.
“We have to reset that and make the assumption that we should only bring a system like that into the world, that we are sure we can control and operates in a subordinate way to us,” he said.
“These tools, like any other past technology, are designed to enhance human wellbeing and serve humanity, not exceed humanity.”