Anthropic to Allow Enterprise Customers to Store Data Logs on Their Own Cloud
Anthropic plans to update its data retention policy, allowing enterprise customers to store their 30-day data logs on their own cloud infrastructure instead of Anthropic's servers. This new security feature is expected to launch later this year to address customer privacy concerns. This policy shift helps Anthropic remain competitive against rivals like OpenAI and Google, which offer zero data retention options. It addresses critical data sovereignty and privacy requirements, which are major hurdles for enterprise adoption of large language models. While Anthropic previously mandated a 30-day retention period on its servers to monitor and investigate potential cyberattacks, the new Bring Your Own Cloud (BYOC) approach mitigates privacy risks while keeping security logs intact. Anthropic has reiterated that it does not use customer data to train its AI models.
## BACKGROUND
In June, Anthropic introduced a policy to retain frontier model data traffic for 30 days to detect and trace cyberattacks leveraging its technology. However, many enterprises operate under strict regulatory frameworks that prohibit sharing sensitive data with third-party AI vendors. To address this, deployment models like Bring Your Own Cloud (BYOC) allow organizations to host vendor software within their own secure cloud environments.