Microsoft CEO Satya Nadella has said AI companies should not ignore serious problems while building powerful AI systems. Instead, they should stop development when they find an issue and fix it before moving ahead. Speaking at the All-In Summit on Tuesday, Nadella said AI should serve humanity and remain under human control. He also stressed the need for user choice, competition and different AI business models. His comments come as AI companies face growing pressure to address safety risks.
Nadella said companies building advanced AI should take enough time to test their models. He also suggested that they should allow independent third-party testers to examine these systems, reports Business Insider.
Also read: Meta One subscription for Facebook, Instagram and WhatsApp is here: Check price, features and more
“It’s possible that they see stuff which are showstoppers before the rest, and if you see a showstopper, stop the show to fix the bugs,” Nadella said.
He compared this approach with software engineering, where developers have to decide how to handle serious bugs. Nadella said companies should not simply ignore an issue because it appears to be an edge case.
“One of the biggest things you learn as an early engineering lead is how to deal with a showstopper bug. Do you stop and fix, or do you defer, or do you go in and say, ‘Hey, this is such an edge case?'” he said.
Also read: OpenAI acquires smartphone camera maker Glass Imaging founded by former Apple employees: Report
Nadella also argued that people should remain in control of AI systems. “We should do what it takes to build stuff that serves humanity first and is in human control,” he said. “You’ve got to have choice. You have to have competition. You have to have all kinds of business models, whether they’re open weights, closed weights, what have you.”
He also raised concerns around privacy and intellectual property. Nadella said users should be able to control how their information is used and should not have to worry about their private knowledge or intellectual property being exposed.
“I want my privacy. I want to be able to embed my knowledge in a set of weights I control. My IP shouldn’t leak,” he said.
His comments come amid a wider debate about AI safety. Some AI leaders have called for more time and stronger safeguards as models become more capable. Others have argued that slowing AI development is not realistic.
Nadella said testing should remain a priority. “I love this idea of having third-party testers,” he said. “We should take all the time we want to test things.”