Hajra Naz
Microsoft has joined forces with Elon Musk’s AI company, xAI, to host its chatbot, Grok, on Microsoft’s cloud platform, Azure. This partnership comes just after Grok stirred controversy with politically charged and misleading responses online.
A Push for Transparency and Reliability
Speaking at a Microsoft-hosted event, Musk said the goal of xAI is to “aspire to truth with minimal error,” while also acknowledging that mistakes will happen. He underlined that building AI that is based on reality is crucial.
In a recorded conversation with Microsoft CEO Satya Nadella, Musk added that Grok will always admit when it makes mistakes. The move to Microsoft’s cloud is expected to make the chatbot more stable, transparent, and easier to manage.
Controversy Sparks Change
The recent backlash began when Grok gave politically sensitive answers that echoed conspiracy theories, particularly about the supposed oppression of white South Africans — a topic often shared by Musk on social media. These responses were not triggered by related prompts, which made the issue more alarming.
xAI later revealed that the incident was caused by an unauthorized code change, violating the company’s internal policies. While the person behind the change was not named, the company committed to several improvements, including
-
Publishing Grok’s system prompts to the public
-
Overhauling its internal review process
-
Setting up a 24/7 monitoring team to catch future issues
Musk didn’t directly refer to the controversy during the Microsoft event, but emphasized that xAI would remain open about any errors made by Grok.
A Subtle Contrast with OpenAI
Some interpreted Musk’s comments as a subtly critical assessment of OpenAI, the business he co-founded in 2015 but later separated from. OpenAI, which now powers many of Microsoft’s internal tools like Copilot, has often been accused of being too secretive about its models and internal workings.
Recently, OpenAI had to fix a glitch that caused its chatbot to give overly flattering responses, raising concerns about its system prompts. In contrast, xAI is leaning into transparency, aligning more closely with open-source approaches like Meta’s LLaMA or DeepSeek’s models from China.
Grok Joins Azure’s AI Foundry
Under this new partnership, Grok will be part of Microsoft’s Azure AI Foundry, a platform offering developers access to a wide range of AI models. It already features tools from OpenAI, Meta, Mistral, Stability AI, DeepSeek, and now, xAI.
At the event, Microsoft’s Nadella described how AI tools are evolving to become more like “virtual teammates” for developers. Tools like GitHub Copilot, used by over 15 million developers, are changing how people write and debug code.
OpenAI CEO Sam Altman, who also appeared at the event via live video, echoed this idea, calling it “one of the biggest changes to programming I’ve ever seen.”
“You now have a real, virtual teammate you can assign tasks to,” Altman said.
Microsoft’s Bigger AI Vision
This collaboration also fits with Microsoft’s overarching plan to employ AI to reduce redundancies and streamline operations. Just last week, the company announced layoffs affecting roughly 6,000 jobs, about 3% of its global workforce, as part of a shift toward AI-driven efficiency.
While the exact number of job cuts wasn’t officially confirmed, Microsoft made it clear that AI will play a bigger role in how the company moves forward.
