Microsoft Adds Grok to Copilot in Word, Excel and PowerPoint
Microsoft is bringing SpaceXAI’s Grok into Copilot, giving some business users another AI model to choose from inside Word, Excel and PowerPoint.
The company announced on Saturday that Grok models from SpaceXAI are being added to Copilot through its Frontier Program. The rollout starts as a preview and is meant to give Microsoft feedback on how customers use the models for everyday work. For now, Grok will only be available to eligible Frontier customers. It will not be offered during the preview in the European Union, European Free Trade Association or the United Kingdom.
Microsoft has also added SpaceXAI to its Online Services Subprocessor List. Companies that want to use Grok have to turn on a separate admin setting, which is disabled by default. Microsoft says administrators will keep control over data processed by SpaceXAI models in supported Copilot experiences.
The move is part of Microsoft’s wider multi-model strategy, which has increasingly focused on letting Copilot customers choose between different AI models instead of relying on a single provider. So now you can choose from GPT, Claude and Grok.
Microsoft says it will look at customer feedback, usage and overall product fit before deciding whether to bring Grok to more Copilot experiences. The company also says models are put through testing, safety validation and Responsible AI reviews before being made available to customers.
Earlier this month, Elon Musk said Grok 4.7 would be coming out this week, but just yesterday he delayed the launch, saying, “Grok 4.7 needs a few more days to cook. We might have penalized response length too much (or something) in RL, as it still gives up on hard tasks (that it can do!) too early and isn’t yet sufficiently rigorous in checking its work.”
Grok 4.7 is expected to be SpaceXAI’s next major model, reportedly scaling up to 2.1 trillion parameters and receiving additional training on internal SpaceX engineering data. Reports say it will focus on better efficiency and stronger performance on complex technical tasks, while potentially trading some response speed for using fewer tokens to complete a job.
Want to see more of our stories on Google?
P.S. Want to keep this site truly independent? Support us by buying us a beer, treating us to a coffee, or shopping through Amazon here. Links in this post are affiliate links, so we earn a tiny commission at no charge to you. Thanks for supporting independent Canadian media!
