Understanding the SurePath AI Public Service Catalog

SurePath AI continuously discovers and catalogs publicly available Generative AI tools on the Internet. Once a tool is identified, relevant metadata is collected and the tool is added to the user interface. As an administrator, you can then choose to allow or deny user access to each new tool.

The collected information is designed to support informed decision-making about user interaction with these tools. This document outlines each available data point to help you evaluate risk, control access, and enforce policy effectively.

Included Data Points

The following data points are provided for each listed service to assist in evaluation, governance, and policy enforcement.

The full list can be found at: https://admin.surepath.ai/config/public-services

Status

Indicates the current availability of the tool and the level of support provided by SurePath AI.

ACTIVE – The tool is currently available on the Internet, and SurePath AI actively supports it.
DEPRECATED – The tool is no longer reliably accessible or is being phased out. SurePath AI currently provides limited support but plans to remove it in the future.

Risk Level

Indicates the potential risk associated with this GenAI tool, based on the criteria outlined.

LOW – Minimal risk identified; tool appears safe and responsible under current usage guidelines.
MEDIUM – Some concerns present; tool may pose moderate risk depending on use case or implementation.
HIGH – Significant risk identified; tool may lead to misuse, misinformation, or security/privacy issues.

Risk Confidence

Represents the degree of certainty behind the assigned risk level above.

LOW – Limited data or weak indicators; risk assessment is uncertain.
MEDIUM – Moderate evidence supports the risk level; some uncertainty remains.
HIGH – Strong evidence or clear indicators support the risk level with high confidence.

Trains on User Data

Many GenAI tools retain user input and may use it to train future models or for other internal purposes.

No – User input is not retained or used for training. (Generally considered safer from a privacy and security standpoint.)
Yes – User input is retained and may be used to improve models or for other uses. (Raises privacy and data governance concerns.)

Some tools offer subscription tiers (often paid) that limit or eliminate data retention and usage for training purposes.

Unmoderated

Indicates whether the GenAI tool applies moderation or filtering to user inputs and/or generated outputs.

No – The tool implements moderation to restrict certain inputs or outputs. (May limit flexibility or transparency, but can enhance safety and compliance.)
Yes – The tool does not moderate inputs or outputs. (Allows maximum flexibility but may increase the risk of harmful or unfiltered content.)

Country

This hosting country for the tool would govern the high level rules that apply to user information and governance of the information collected by the tool itself.

SurePath AI Support

Access Control – Enables administrators to control access to the tool. This setting is available as an on/off toggle in both Default and Group policies.
Request Intercept – When available, the tool can capture, log, and modify the user’s input (prompt) for auditing or analysis. This feature support is required to ensure that sensitive data is not sent to the tool.
Response Intercept – When available, the tool can capture and log the AI-generated response returned to the user.

Note: Access Control applies universally to all tools in the catalog, regardless of whether it is surfaced in the tool’s user interface.

Company

The organization responsible for developing, maintaining, and operating the tool.

Description

This section provides an overview of the tool, including its intended purpose, authorized use cases, any restrictions or limitations on usage, and other relevant operational or contextual information necessary for appropriate use and governance.