Microsoft's Nadella warns against concentration of AI power
Remarks come as Microsoft signs a 20-year agreement with Chevron to power a new giant datacentre in Texas
Microsoft chief executive Satya Nadella has warned that the future of AI cannot be controlled by a small number of companies, arguing that the industry must demonstrate broader economic and social benefits if it hopes to maintain public support.
In an interview with The Wall Street Journal, Nadella delivered a strong critique of the direction of the AI race.
He questioned the current model in which a handful of firms dominate the development of the world's most advanced systems while also warning of large-scale job disruption and safety risks.
"You can't say, hey, all white-collar jobs are gone and this could even be a weapon and we will use all the power to build datacentres," he said.
While Nadella did not mention specific rivals by name, his remarks appeared aimed at companies leading the development of frontier AI models, including OpenAI (in which Microsoft has invested billions), Anthropic and Google.
He argued that the public would be unlikely to accept a future in which only a small group of companies controlled the technology's development and deployment.
"The public won't tolerate just a few models and companies doing all of the learning for the world," he said.
Push for cheaper AI
Microsoft has invested billions of dollars in AI partnerships, most prominently with OpenAI, helping fuel the rapid expansion of genAI over the past several years.
More recently, it has introduced a range of lower-cost AI models and tools designed to reduce costs for customers as businesses grapple with rising spending on AI services.
Microsoft has also launched new features that allow customers to choose between different AI models for long-running tasks, rather than relying on a single provider.
Nadella's stated vision centres on a more competitive marketplace where organisations can combine different models with their own data and expertise, reducing dependence on a handful of AI suppliers.
Microsoft is also reportedly considering hosting versions of low-cost AI models developed by Chinese start-up DeepSeek. Such a move could intensify competition in a market already facing growing pressure on pricing.
Debate over jobs
Nadella challenged predictions that AI will inevitably lead to widespread job losses.
Anthropic chief executive Dario Amodei has previously suggested AI could eliminate a large share of entry-level office jobs within a few years, while OpenAI's Sam Altman has warned of significant workplace disruption.
Nadella argued that businesses should focus on redesigning jobs rather than simply replacing workers.
"No, how about we think about reorganising the jobs?" he said.
He said companies would need to combine human expertise with AI capabilities, describing future organisations as "continuous learning systems" where workers and AI tools operate together.
According to Nadella, AI firms must provide a clearer roadmap showing how businesses can manage the transition, acknowledging that technological change would inevitably involve disruption but insisting there was a practical path forward.
Restoring confidence in AI
Nadella argued that restoring confidence in AI would require more than reassuring rhetoric.
The industry, he said, must demonstrate tangible benefits and ensure people feel they retain control over their futures.
"No amount of just narrative is going to do it because where we are now, we have to sort of walk the walk," he said, adding that AI companies must create genuine opportunities for people to participate in and benefit from the economic changes the technology brings.
"We now have to do the hard work in earning the social permission."
Nadella’s remarks came as the company signed a 20-year deal with fossil fuels giant Chevron to develop a hyperscale datacentre in Texas, codenamed Kilby, that’s expected to deliver approximately 2.67 GW of capacity.