Microsoft has announced that it will add Anthropic AI models to Microsoft 365 Copilot within Word documents by late May. This will help you tidy up documents with the assistance of artificial intelligence. The models will be enabled by default outside of the EU, EFTA, and UK. Inside those regions, it will be disabled by default, and can be enabled by the admin.
To use this feature, you must be a Microsoft 365 tenant with Microsoft 365 Copilot (Premium). You’ll find the model in addition to OpenAI models from the model selector in Copilot in Word.
For admins in the EU, EFTA, or UK, there will be an option to enable or disable Anthropic at the tenant level through the Microsoft 365 admin center. While Microsoft probably dislikes that it has to do this, it’s actually good for organizations in those regions because nothing gets changed in this regard without express permission.
To change the availability of the Anthropic models, go to the Microsoft 365 admin center and go to Copilot > Settings > View all. From there, select ‘AI providers operating as Microsoft subprocessors’, under ‘Available subprocessors for your organization’ choose ‘Anthropic’ and select ‘Enable’ or ‘Disable’.
Anthropic models will be available in Word as an AI provider operating as a Microsoft subprocessor and all existing Microsoft 365 Copilot security, compliance, and governance controls will continue to apply. If the models are left disabled then there will be no change in the user experience.
The Redmond giant recommends telling help desks about the change and updating internal documentation. However, with just two weeks to go until the feature arrives, that leaves very little time to follow these recommendations.
It will be interesting now to see whether Microsoft brings in other AI models from other companies such as Google and Meta to give users more choice. You can find out more about this change under Roadmap ID 558440 or Message ID MC1296489.