Understand how Foxit protects data across AI-powered document workflows
Foxit AI Assistant brings AI-powered document chat, summarization, rewriting, translation, content creation, and other intelligent capabilities to Foxit PDF Editor and Foxit Workspace. This security overview explains the architecture, data-handling practices, and controls that support these experiences, helping security, privacy, IT, and compliance teams evaluate how information is processed across different workflows.
Foxit AI Assistant uses processing paths that vary by product, feature, platform, configuration, and entitlement. Eligible PDF Editor workflows may use Foxit-managed cloud processing, On-Device processing with Local SLMs, or Custom processing through a customer-configured compatible endpoint. Foxit Workspace operates through cloud-connected services and does not support On-Device processing.
What the overview covers
Clear processing boundaries: See how content moves through Foxit-managed cloud services, supported on-device workflows, customer-configured model endpoints, and authorized third-party tools.
Purpose-limited use of customer content: Customer documents, prompts, AI outputs, and other customer content are used to provide the requested capability. Foxit does not use—or authorize or instruct subprocessors to use—this content to train or fine-tune AI or machine-learning models unless the customer expressly authorizes it.
Data lifecycle and regional processing: Understand how request data, locally processed data, Workspace Project content, temporary working data, and operational records are handled and retained. The overview also describes Foxit’s regional cloud architecture and considerations for international transfers.
Security and access controls: Learn about encrypted network transport, authentication, authorization, logical access boundaries, data minimization, retention controls, and safeguards for stored product data.
User control over tools and actions: Review how Connectors are configured, how authorized scopes and credentials limit access, and how selected sensitive or consequential operations can require user approval.
Defined responsibility boundaries: Understand the respective responsibilities of Foxit, customers, and third-party providers for cloud, on-device, Custom endpoint, and Connector workflows.
Human review of AI output: AI-generated results can be incomplete or inaccurate. Users should verify material outputs against source documents and other authoritative information, especially in high-impact contexts.
Foxit’s AI Assistant is designed to provide enhanced document productivity while safeguarding user privacy and adhering to the highest security standards.
Download the full whitepaper to learn how Foxit AI Assistant ensures document security while improving workflow efficiency.