Microsoft reviews data retention rules before broader internal Claude use, according to report

Microsoft is said to be putting guardrails around its internal use of Anthropic’s Claude, according to a market report citing a review of Anthropic’s data retention practices.

The report says Microsoft is assessing Anthropic’s approach to how customer or employee inputs are stored and retained before it moves beyond its current, narrower internal use.

The need for review highlights the compliance and risk-management friction that can arise when companies deploy third-party AI systems, especially in environments where internal data handling policies are tightly controlled.

Microsoft is also under broader scrutiny from customers and regulators on how it manages data in AI workflows, including questions around what data is used for model improvement, how long it is kept, and who can access it. In that context, even small differences in retention policies can change how quickly a tool can be scaled.

The report frames the data retention review as a precondition for “wider internal use,” indicating that Microsoft’s rollout is contingent on aligning Anthropic’s retention rules with Microsoft’s internal governance requirements.

A key point not spelled out in the report is what exactly Microsoft plans to change or demand. It does not provide details on whether Microsoft will require shorter retention periods, stronger isolation, additional controls for enterprise users, or limits on how prompts and outputs are stored.

It also does not state whether the review affects any specific business units, whether there is a temporary pause on certain AI tasks, or what timeline Microsoft is working toward for a broader rollout.

For now, the most observable takeaway is that Microsoft’s AI operating model is increasingly shaped not only by model capability, but also by the legal and operational specifics of data retention and downstream handling.