Microsoft CEO Satya Nadella has defended the term “superintelligence” as a way to describe artificial intelligence, while warning that increasingly capable models also pose security risks if companies fail to establish adequate safeguards.
“You’re welcome to talk to the president of the United States. He named it. And I mean, it’s as good a name as anything as far as I’m concerned,” Nadella told Reed Albergotti on Semafor Technology’s “Future Proof,” discussing the renaming of AI to SI.
READ: Microsoft’s Satya Nadella says companies are ‘paying for their own IP,’ offers fix (July 13, 2026)
In an interview published on Oct. 21, Nadella discussed the challenges of AI alignment, the difficulty of understanding how advanced models operate, and Microsoft’s approach to securing AI systems as their capabilities continue to evolve.
Nadella acknowledged that the industry has made significant progress in developing more capable AI models but said its ability to ensure their behavior remains safe and predictable has not kept pace.
“The good news here is the scaling laws have worked and worked magically a lot better and a lot more than I would have ever imagined in 2018,” he said.
However, he described scaling laws as “more of an empirical finding that has stood the test of time, like Moore’s Law,” adding that “alignment seems elusive.” Nadella said he hoped researchers would eventually find solutions to these challenges.
He said companies should not rely exclusively on safeguards built into AI models. Instead, they need additional layers of security, monitoring and control outside the models themselves.
He compared advanced AI systems to employees within an organization who have access to sensitive information and resources but could potentially misuse that access.
“You have effectively these intelligences that are inside that have all the capability to be an insider risk,” he said.
Nadella said Microsoft’s approach involves establishing control and assurance mechanisms outside the model, using multiple models and making systems less opaque. The approach is intended to help organizations deploy AI while maintaining stronger oversight.
He also discussed Microsoft’s security initiative, including Agent 365, which is designed to help organizations manage and govern AI agents.
READ: Microsoft CEO Satya Nadella warns of few AI giants ‘destroying entire industries’ (June 15, 2026)
Asked about the commoditization of AI models and the industry’s shift from measuring costs per token to costs per task, Nadella said, “At the end of the day, everything is a commodity and nothing at scale is a commodity.”
He went on to explain that cloud computing, too, is a commodity used to deliver a variety of services.
Asked whether he had anyone in mind to succeed him as CEO, Nadella said Microsoft has several capable, ambitious and talented candidates. They include both executives who have risen through the company’s ranks and those who joined from outside and are now performing well.