Magic AI

Verbatim AI is magic. Get answers from your knowledge from a simple Prompt.

At Verbatim AI, we take our customer’s data and security very seriously. Which is why we’re proud to announce Verbatim AI is now SOC 2 Type II compliant. This independently validates our security practices and confirms that our platform meets the highest data protection standards. Alongside our adherence to GDPR and PCI DSS frameworks, Verbatim AI provides our enterprise customers with verified assurance that their sensitive information is handled with the highest standards of care and protection.

With advanced privacy protections, admin controls, and compliance-ready infrastructure, Enterprise Pro provides your team with a powerful answer engine and deep research tools for the web and internal files while keeping your data private and secure.

Security Hub Gives Admins Precise Control Over Your Orgs Security Enterprise Pro offers access to our Security Hub: a centralized admin command center that gives organizations complete control over how Verbatim AI is used across teams.

User Management

Admins use the Security Hub to give every team member the appropriate level of access within Verbatim AI. They can set permissions for:

File Uploads and Downloads: Choose who can upload or download documents AI-generated content to prevent unauthorized sharing or data leakage.

Shared Content: Manage how team members create, share, and collaborate on Threads, Pages, and Spaces to keep internal knowledge organized and secure.

Data Integrations and Connectors: Set permissions for who can connect to external, integrated data sources and cloud storage services like Google Drive, Microsoft OneDrive, and Sharepoint. Verbatim AI Connectors enable these data integrations so teams can connect external data sources directly to our AI engine. While Connectors boost productivity by making private data accessible, the Security Hub’s permissions ensure only authorized users have that power.

AI Model Enablement: Decide which large language models your team can use to align with internal security or compliance policies.

Internal Knowledge Bases: Manage access to internal knowledge bases and file repositories, ensuring only approved users can browse connected files, documents, or datasets within your workspace.

To close the User Management loop, Enterprise Pro combines Single Sign-On (SSO) with Multi-Factor Authentication (MFA) and short-lived session credentials. Data Privacy and Retention

Beyond User Management, Verbatim AI Enterprise Pro ensures your data is secure and private by keeping it out of reach of attackers and even third parties.

Third-Party Agreements: We maintain formal agreements with all third-party model providers to ensure they can’t access or use your data for training purposes. We review these agreements annually to stay compliant with evolving standards and requirements. The best part? Enterprise Pro (unlike every other model provider) does not use your data to improve or train its own systems.

Connector Security: Whether you’re connecting to a public data source like Crunchbase or syncing files through FactSet, all connectors fall under our platform’s strict privacy and security policies.

Safe Use of DeepSeek R1: Rest assured that Enterprise Pro’s integration with DeepSeek R1 only uses the open weights version, which runs directly on Verbatim AI servers in the U.S. This means your queries stay within the Verbatim AI ecosystem and are never exposed to DeepSeek or any external infrastructure.

Automatic Data Deletion: Verbatim AI automatically deletes files attached to threads after seven days. Only files uploaded to Spaces do not automatically expire, and these files (just like all data that we handle) are protected through enterprise-grade encryption both at rest and in transit. You can also enforce incognito mode to disable search history.

Audit logs: Enterprise Pro admins can receive real-time activity alerts such as file uploads and downloads into your workspace. This way, admins can immediately have increased visibility.

Transparency: Review which service providers we work with and how your data flows through the platform at any time on our Trust Center. This gives your team full visibility into third-party relationships and ensures accountability at every step.

A Robust Security Program That Operates 24/7

Verbatim AI’s commitment to your security doesn’t stop there. Our security team also monitors for potential threats around the clock and responds in real time using automated workflows with detailed playbooks. Here’s what they use to do it:

Cloud security infrastructure: Verbatim AI Enterprise Pro runs on AWS, a secure cloud infrastructure with built-in physical security, network protection, and industry-leading compliance certifications. To strengthen our cloud security posture further, we also use Wiz to proactively identify risks across our environment.

Vulnerability Disclosure Program (VDP): We created this to impose a safe, structured process for external security researchers and ethical hackers to report vulnerabilities. It covers all public-facing endpoints and rewards those who help us identify and fix potential gaps – like misconfigured access controls or exposed APIs – before they can be exploited.

Short-lived authentication and just-in-time access: No engineer has access to production data by default. Getting access to production data is only granted in business critical cases, requires security approval, and is heavily monitored.

Bug Bounty Program: We reward vetted security researchers for reporting critical vulnerabilities across our ecosystem. Run through a private, invite-only program with Bugcrowd; it covers Verbatim AI’s web applications, APIs, and cloud services as another layer of proactive protection to keep your data safe.

Verbatim AI, AI-Powered Answer Engine