## Related terms - workspace - user - profile - synerise\-platform See the [glossary](/glossary.md) for definitions. --- # System Agent FAQ
This feature is available in [public preview](/docs/log-in-and-navigate-synerise-platform#public-preview) mode.
This article answers frequently asked questions about the regulatory classification of the [System Agent](/docs/ai-hub/synerise-system-agent/introduction-to-the-system-agent), how it processes data, and how you can control it. In this article, "Customer" means the organization that uses the Synerise Platform under an agreement with Synerise. ## Regulatory classification --- ### Is the System Agent an AI system under the EU AI Act? Yes. The System Agent is considered an AI system within the meaning of the EU AI Act. It is designed to operate with varying levels of autonomy and uses AI models to generate outputs and support users in performing tasks within the Synerise Platform. Based on its intended purpose, the System Agent is not classified by Synerise as a high-risk AI system under the EU AI Act. In particular, the System Agent is not intended to be used for any of the high-risk purposes referred to in Article 6 and Annex III of the AI Act. Customers remain responsible for assessing their specific use of the System Agent and ensuring that it is not deployed for purposes that could result in a different regulatory classification or trigger additional obligations under the AI Act. ### Is the System Agent a general-purpose AI model? No. The System Agent is designed as a domain-specific AI system supporting users in performing tasks within the Synerise Platform. It may integrate or rely on general-purpose AI models provided by third parties. The use of a general-purpose AI model as an underlying component does not in itself make the System Agent a general-purpose AI model. ## Data processing --- ### Does the System Agent process the Customer's business or personal data, including profiles? Potentially, yes. The System Agent may process the Customer's business and personal data available in the workspace, depending on the [System Agent configuration](/docs/settings/agentic-ai/system-agent) applied by the Customer and the Customer's use of the System Agent. ### Where is Customer data processed when using the System Agent? The System Agent relies on AI models provided through the infrastructure of third-party cloud providers that are already involved in processing the relevant Customer data as part of the Synerise Platform services. Therefore, the use of the System Agent does not require Customer data to be disclosed to an additional or new third-party provider. The processing of personal data in connection with the System Agent is covered by and subject to the Data Processing Agreement (DPA) between the Customer and Synerise. The applicable data processing location remains subject to the infrastructure and regional configuration used for the relevant Customer and the applicable cloud services. ### Is Customer data used to train AI models? Customer data, including prompts, outputs, and data accessed by the System Agent, is not used to train models for the benefit of other Synerise customers or third parties (including AI model providers). ## Control and reliability --- ### Does the System Agent respect existing user permissions? Yes. The System Agent operates within the context of the Customer's workspace and the permissions applicable to the relevant user. What the agent can access in a workspace depends on the [System Agent settings](/docs/settings/agentic-ai/system-agent), and what it can do for a particular user depends on that user's [permissions](/docs/settings/identity-access-management/permissions/ai-hub-permissions#use-the-system-agent-with-workspace-resources). ### Can the System Agent provide incorrect answers? Yes. As with other AI-assisted systems, outputs may contain inaccuracies or may not fully reflect the user's intended context. Users should review outputs before relying on them, particularly where they may affect customers or business decisions. For more information, see the [Legal notice](/docs/ai-hub/synerise-system-agent/introduction-to-the-system-agent#legal-notice). ### Can the System Agent make decisions about individuals? The System Agent is designed as an assistive tool for Platform users. Whether a particular Customer configuration results in automated decision-making concerning individuals depends on the Customer's use case, configuration, and subsequent use of Synerise Platform functionalities. ## Agreement and usage --- ### Is the System Agent covered by the existing Synerise Agreement? Unless otherwise agreed, the System Agent is provided as part of the Synerise Platform and is subject to the existing Agreement, including applicable Terms and Conditions, limitations of liability, and other contractual provisions. Additional terms may apply to this functionality. Please note that the disclaimers indicated in the [Legal notice](/docs/ai-hub/synerise-system-agent/introduction-to-the-system-agent#legal-notice) also apply to the System Agent. ### Is the System Agent subject to additional fees? No, but its usage is subject to the limitations described below. ### Are there any usage limits or restrictions applicable to the System Agent? Yes. Access to the System Agent may be subject to technical, operational, or commercial limitations, including usage quotas, rate limits, and restrictions on certain operations, as specified in [Usage limits](/docs/ai-hub/synerise-system-agent/introduction-to-the-system-agent#usage-limits). ### Who is responsible for AI-generated outputs and actions performed through the System Agent? The Customer is responsible for reviewing and verifying AI-generated outputs before relying on them. Synerise does not guarantee the accuracy or completeness of such outputs. ## Related articles --- - [Introduction to the System Agent](/docs/ai-hub/synerise-system-agent/introduction-to-the-system-agent) - [Enabling System Agent](/docs/settings/agentic-ai/system-agent) - [Synerise MCP Server FAQ](/developers/mcp/faq)