Microsoft adds Anthropic’s AI to Copilot (techcrunch.com)

Microsoft announced it will add Anthropic’s Claude models to Copilot, letting business users pick between OpenAI-powered options and Anthropic’s Claude Opus 4.1 and Claude Sonnet 4 for specific tasks. The move follows a recent agreement to integrate Anthropic into Office 365 apps and marks another step away from Microsoft’s prior near-exclusive reliance on OpenAI. In Copilot, Opus 4.1 will be recommended for complex reasoning, architecture planning and advanced coding, while Sonnet 4 targets routine development, large-scale data processing and content generation—so users can route queries to models optimized for different types of work. For the AI/ML community this matters for both product and enterprise strategy: it formalizes multi-vendor model selection inside a major productivity assistant, increasing redundancy, reducing vendor lock-in and letting organizations match workloads to model strengths (and likely cost/performance tradeoffs). Technically, it implies richer model-selection and routing logic in Copilot, new integration surfaces for building enterprise-grade agents and custom tools, and comparability between high-end reasoning models across vendors. It also intensifies competition around model capabilities, safety behavior, and deployment options for large-scale, mission-critical AI in the workplace.