As announced in MC1176368 (October ’25) GPT-5 is now the default model and today we’re introducing the ability for your users to select how much reasoning to apply.
Users can choose from three modes:
- Auto (default): Copilot decides how long to think.
- Quick response (chat): Copilot answers right away.
- Think deeper (reasoning): Copilot thinks longer for better answers.
This change is associated with Microsoft 365 Roadmap ID: 532736
General Availability (Worldwide): Rollout will begin December 4, 2025, and we expect to complete by December 8, 2025.
[How this affects your organization:]
Who is affected: All users of Microsoft 365 Copilot.
What will happen:
- Users will see a new selector in Copilot experiences.
- Default mode is Auto; users can switch to Quick response or Think deeper.
- User’s choice is remembered and retained for new chats.
- No admin configuration changes required. There are no admin controls for this feature.
- Feature is enabled by default for all tenants.
- No action required; this is a standard rollout.
- Optional: Communicate the new modes to helpdesk staff and users.
No compliance impacts identified. Review your organization’s policies as needed.
Source: Microsoft
Latest Posts
- Cloudflare Tunnel, Cloudflare Tunnel for SASE, Cloudflare Mesh, Gateway, Cloudflare One – Hostname routing is now generally available, with a new public IP range for initial resolved IPs
- langcache-embed-v3-small, Mellum2-12B-A2.5B-Thinking, and LightOnOCR-2-1B models now available on Amazon SageMaker JumpStart
- Amazon EC2 High Memory U7i instances now available in AWS South America (São Paulo) region
- GLM-5.2 FP8, NVIDIA-Nemotron-Nano-12B-v2 and GLM-OCR models now available on Amazon SageMaker JumpStart
