Every story tagged Data Retention, curated for CIOs and IT leaders — ranked by source credibility, engagement, and freshness.
4 stories · open in the command center
Claude Code automatically deletes local context history after 30 days by default across all user tiers, with commercial users (Team, Enterprise, API) having standard 30-day retention and consumer users opting into either 5-year or 30-day retention based on model training preferences. IT leaders should be aware that while Anthropic doesn't train on commercial code without explicit opt-in, session transcripts are retained on Anthropic servers during remote sessions for device synchronization, and organizations can implement zero data retention policies or disable feedback collection through configuration controls. This represents a significant privacy and compliance consideration for enterprises evaluating Claude Code deployment, particularly regarding code exposure, audit trails, and data governance requirements.
Anthropic is implementing mandatory 30-day data retention for its advanced Mythos and Fable models (effective June 9, 2026) to detect sophisticated multi-request attack patterns, affecting only organizations currently using zero data retention across API, AWS Bedrock, Google Cloud, and Microsoft platforms. This policy introduces a compliance burden for enterprise IT organizations that must choose between enabling data retention to access next-generation AI capabilities or maintaining zero-retention postures, while Anthropic implements technical safeguards including tamper-proof access logs, restricted reviewer access, and optional customer-managed encryption. Technology leaders must evaluate the security and privacy trade-offs of enhanced AI capabilities against organizational data governance policies and determine whether the 30-day retention window and Anthropic's security controls meet their regulatory and operational requirements.
Google is expanding its data collection practices by automatically saving user interactions across Google Lens, Search Live, voice searches, and Translate for AI model training and personalization, creating significant privacy and compliance implications for enterprises managing corporate data use. This change separates from existing privacy controls, requiring IT organizations to reassess data governance policies and employee awareness programs, particularly regarding how user-generated content feeds AI development. CIOs must understand the new 'Search Services History' setting and its business implications, as this represents a broader industry trend of leveraging user interactions for AI advancement that may conflict with organizational data retention and privacy policies.
Microsoft has restricted internal employee access to Anthropic's new Claude Fable 5 model due to data retention requirements that conflict with Microsoft's zero-data-retention policies, creating legal compliance concerns around customer data and confidential information handling. This restriction signals that enterprises must carefully evaluate third-party AI models' data governance practices before adoption, as safety-focused architectures may introduce unacceptable data residency and retention risks. The situation highlights a critical tension between AI capability advancement and enterprise security/compliance requirements that will likely affect broader AI procurement decisions across the industry.