The Business, Standard and Plus editions of Gemini Enterprise are built on Google’s secure-by-design infrastructure, giving you the peace of mind to deploy AI agents across your organisation confidently.
- You own your data, not Google
- Your data – including prompts, outputs, and training – aren’t used to train Google models or models for any other customer
- Google never sells customer data to third parties
- Google doesn’t use customer data for advertising
The Standard and Plus editions of Gemini Enterprise meet strict cloud compliance requirements around retention, accessibility, and maintaining integrity of your records, such as:
- Data residency to help ensure your data stays in the region of your choice
- Google Cloud audit logging to preserve immutable electronic logs on your company’s activity and help your security, auditing, and compliance entities monitor for possible vulnerabilities or external data misuse
- Customer Managed Encryption Keys (CMEK) through Cloud Key Management Service to encrypt your data with customer-controlled keys, giving your organization control over protection level, location, rotation schedule, usage and access permissions, and cryptographic boundaries
- VPC Security Controls to block access from untrusted locations outside your organization’s network and mitigate data exfiltration risks
- Industry-standard compliance
The Standard and Plus editions of Gemini Enterprise support various identity providers to authenticate users and enforce data access controls from connected enterprise systems. It verifies both user access to the application and access to data from connectors to ensure that content within Gemini Enterprise is accessible only to authorised individuals.
Gemini Enterprise enforces identity and access controls at various levels:
- Cloud Identity and Access Management to restrict access to provisioned admins and users
- The ability to select the content indexed from connectors and use the minimum level of permissions required to index data
- Support for Google and third-party identity systems and user groups to authenticate users to Gemini Enterprise
- Enforcement of document-level access controls to help prevent potential leakage of information via search, assistant, or agents
All editions of Gemini Enterprise include Model Armor, our built-in guardrails for AI prompts and responses, to proactively screen malicious and unsafe interactions such as prompt injection, jailbreaking, and sensitive data detection.