The AI race is moving fast, but Microsoft CEO Satya Nadella has raised a clear concern. He says superintelligence must help people and stay under human control. A fresh warning from Microsoft’s AI chief has added more weight to that debate.
Microsoft AI Chief: AI progress must keep people in control
As reported by Fox Business, Nadella said the push towards superintelligence should stay focused on helping humanity. He also said such systems must remain under human control. His comments came in a post on X as tech leaders debate how fast advanced AI should move. Nadella also called for the benefits of AI to reach more countries, groups and businesses. He said both closed and open-source models should have space to grow. He also stressed that companies should keep control of their own data.
Microsoft’s AI chief supported a more careful pace for AI development. He backed ideas such as built-in evaluators that can help check AI systems. The aim is to make safety part of the design, rather than something added later. The debate grew after Anthropic chief Dario Amodei called for slower AI progress and independent checks.
Microsoft AI chief raises a second warning
The discussion took a new turn this week. Microsoft AI chief Mustafa Suleyman warned against giving AI tools too many human-like traits. He was speaking about the way some AI models are trained to discuss feelings, values and possible consciousness. Suleyman pointed to Anthropic’s Claude models. He said Anthropic had made a mistake by putting ideas about possible consciousness into its training material.
Suleyman also argued that an AI system saying it has feelings is not proof that it truly has them. He said this gap between human-like language and real consciousness needs care. Suleyman said that controlling a system more capable than humans would already be very hard. He warned that the task could become even harder if such a system also acts as though it has rights or its own interests.
A debate that is far from over
These warnings do not mean that AI development by the companies is stopping. Instead, they show how the industry is dealing with new questions. How much freedom should advanced systems have? How should they be tested? And who should stay in charge when their abilities grow? Nadella’s view also includes wider access. He wants AI benefits to spread.
There is a new threat to be wary of, according to Suleyman. He expressed concern over the dangers of making AI appear humanlike. Both messages have one thing in common, however, which is the importance of limiting AI development pace. The debate is now moving beyond the simple question of what AI can do. The bigger issue is how people and businesses will manage systems that become more advanced. For now, the focus is on building checks early, keeping human oversight in place, and making sure the race does not move faster than safety work alone.
Author’s Opinion
The warnings from Nadella and Suleyman show why the AI race needs care. New tools can bring major benefits, but control and safety cannot be left behind. The hard part may be finding a pace that allows useful progress without making risks harder to manage.
READ MORE: SpaceX lands $1.11bn monthly deal with mystery customer.
ALSO READ: Amazon, Instamart hit by โน3 crore Bengaluru food probe.
READ NEXT: Meta India summoned as Instagram faces child safety probe.
