Security description for TeamEngine AI Services

TeamEngine's AI Services are designed to support secure, controlled and transparent use of artificial intelligence within the TeamEngine platform.

This page describes the technical safeguards, internal access controls, provider security measures and transfer mechanisms applied when AI functionality is used.

Current AI capabilities

The AI Services are delivered through AI Engine, the assistance layer in the TeamEngine portal. All capabilities share one activation, one set of access controls, one credit balance and the same AI Provider configuration described on this page. This section is updated when capabilities are added or changed.

• Document questions - in Meeting Zone, the meeting and document reader. Users can ask questions about the document they have open and receive answers grounded in its content, for example a summary of a section, where a figure is discussed, or which decision a paper proposes. Each user's conversation is saved for that user only, in the customer's own portal.
• Meeting text and minutes assistance - in the calendar module. Summarize, rewrite and bullet-point helpers for meeting text while minutes and meeting material are being written. The submitted text is used only for the duration of the request, and no conversation history is kept.

Common to all capabilities

• Answers are produced only from the content the user is working in. The AI Services do not browse the internet and cannot reach material the user is not already allowed to open in the portal.
• AI Engine is off until a portal administrator activates it, and the administrator can switch it off at any time.
• Customer input and output are not used to train AI models.
• Output is machine-generated and must be reviewed by the user before it is relied on.

Technical security measures

TeamEngine applies, and requires its AI Providers to apply, technical security measures appropriate for the processing performed within the AI Services.

Logical isolation

Customer data is logically separated from other organizations' data at the AI Provider level.

Encryption in transit

Data transmitted between TeamEngine and AI Providers is protected using TLS 1.2 or later.

Encryption at rest

Data stored by relevant AI Providers is encrypted at rest using AES-256 or an equivalent encryption standard.

Access control

Access to systems and data is restricted according to the principle of least privilege and protected by appropriate authentication controls.

Logging and traceability

AI-related processing events may be logged for operational security, troubleshooting, abuse prevention and traceability.

Security filters

Security filters, abuse detection and content moderation controls are configured where supported by the AI Services.

TeamEngine internal access controls

TeamEngine personnel do not have general access to customer input or output processed through the AI Services.

Where internal access is required for operational, support, security or service improvement purposes, access is restricted and controlled.

• Access is limited to authorized personnel with a specific work-related need.
• Access is restricted according to the need-to-know principle.
• Personnel with access are subject to confidentiality undertakings.
• Internal access is logged where technically applicable.
• Access rights are reviewed and removed when no longer required.

AI Provider security

TeamEngine uses AI Providers that apply recognized security, privacy and cloud governance controls relevant to the processing performed by the AI Services.

Relevant security and privacy certifications may include:

• ISO/IEC 27001 — Information security management
• ISO/IEC 27017 — Cloud-specific security controls
• ISO/IEC 27018 — Protection of personal data in cloud services
• ISO/IEC 27701 — Privacy information management
• SOC 2 Type II

TeamEngine reviews relevant provider documentation as part of its vendor governance and sub-processor management.

Current AI model and processing location

TeamEngine's AI Services currently use the Gemini 3.8 Flash model from Google, accessed through Google Cloud Vertex AI.

AI requests are sent to the Vertex AI EU multi-region, so processing takes place in Google Cloud data centers within the EU.

The model and region may change as the AI Services develop. Changes to AI Providers are handled through TeamEngine's sub-processor management, and this page is updated accordingly.

Data transfers

Where the use of an AI Provider involves a transfer of personal data outside the EU/EEA, TeamEngine applies transfer mechanisms in accordance with applicable data protection law.

These mechanisms may include the EU-US Data Privacy Framework, where applicable, and the European Commission's Standard Contractual Clauses, supplemented by additional safeguards where required.

Governance and monitoring

TeamEngine continuously evaluates the AI Services from a security, privacy and operational perspective.

• AI Providers are assessed before being used in the service.
• Security and privacy documentation is reviewed as part of provider governance.
• Access controls and internal procedures are reviewed on a recurring basis.
• Changes to AI functionality are assessed from a security and data protection perspective.