Microsoft Joins Calls for AI “Pacing,” Pledges to Build Kill Switches Into AI Products

Microsoft on Monday published a new AI “code of conduct” promising to avoid building products that could evade human oversight, following CEO Satya Nadella’s call on Sunday for the tech industry to “pace” its progress in developing new AI.

The measures follow similar calls for pacing from leaders of Anthropic, OpenAI, xAI, Meta, and other tech companies over the weekend. Microsoft said in its AI code of conduct that it would refrain from building AI models whose behind-the-scenes thinking isn’t clear to humans, a growing concern inside other labs like OpenAI.

“If it isn’t safe we shouldn’t build it,” Microsoft AI CEO Mustafa Suleyman said in an interview. “We’re designing this as a training document showing our intention… and this is the driving force between the humanist AI code of conduct, that people matter more than AI, that we should make safety a priority.“

Microsoft is also open to letting third-party auditors that would examine whether its internal AI development adheres to such a slowdown, a policy proposed by Anthropic CEO Dario Amodei on Saturday, according to someone with knowledge of the company’s position.

Microsoft trains its own AI models that compete with those from Anthropic and OpenAI, although the company’s contract with OpenAI barred it from developing highly advanced AI until the deal was amended late last year. Microsoft has more recently put out AI products that use a combination of its own models, along with those from Anthropic and OpenAI, which it says can achieve just as sophisticated cybersecurity tasks as those from the competing labs.