Requirement: FedRAMP Moderate certification is required in order for the City of Seattle to enable the AI Assistants in our Production Web GIS environments. We understand that all 'General Availability' AI Assistants will have FedRAMP Moderate certification.
Problem(s):
ESRI has informed us “Preview is a stage between beta and general availability. Esri states that preview capabilities are documented in product documentation, are part of a Limited-Use Program, and come with limited security and privacy validation. Esri also states that security and privacy validation will be complete at general availability.” Due to the ”limited security and privacy validation”, any AI Assistant in the Preview stage will not meet Seattle IT Security requirements and therefore will not be approved for use at the City.
There is currently no method for only enabling 'General Availability' AI Assistants. Although there are two methods for controlling which AI Assistants are available to users (AI Assistant setting to Allow use of AI assistants by members of your organization; and the Security - Apps - Block Esri apps and capabilities while they are in beta, Enabling both still allow Preview categorized AI Assistants to be available to users.
Requests; We would like Esri to consider the following:
Minimum Viable Option - Create a setting that would allow only 'General Availability' AI Assistants to be available to users, provided ‘General Availability’ also means that all General Availability AI Assistants are FedRAMP Moderate certified.
Preferred Option - Create setting(s) that allow Administrator to selectively enable individual AI Assistants. This would allow us to not only provide AI Assistants that met the FedRAMP Moderate certification but also to ensure the available AI Assistants have proper governance in place and can be supported for general use. We pride ourselves in offering self-service platforms with nearly 3000 users that we support. We would like to have confidence in the capabilities that are enabled so that we can provide and support a positive user experience for all who opt to use these new AI Assistants. Without this option, there is a risk that upon the release of new AI Assistants through Updates/Upgrades we may need to disable use of all AI Assistants while the new ones go through the required IT Security & Privacy reviews and other vetting. This would produce a disruptive and negative experience for our users.
