Learn how to create a policy for Nightfall for ChatGPT from within Nightfall.
You can now configure policies on ChatGPT to record violations and set data that is deemed to be sensitive in nature to your organization.
The policy creation involved the following stages.
Configure Detection Rules for ChatGPT
Learn how you can select the ChatGPT integration in a Nightfall policy.
In this stage, you select the Integration for which the policy is created. In this case, the ChatGPT integration must be selected.
Click Policies from the left menu.
Click + New Policy.
Select Sensitive Data.
Select the ChatGPT integration.
Learn how to configure the advanced settings section in Nightfall policies created for ChatGPT.
This stage allows you to select notification channels if a policy violation occurs. The notification alerts are sent at two levels.
This section allows you to send notifications to Nightfall users. The various alert methods are as follows. You must first turn on the toggle switch to use an alert method.
The alert configurations configured in this section describe the process of creating alerts at the policy level. Policy-level alerts apply only to the policy on which they are configured. To configure an alert on all the Gmail policies, you must configure alerts at the integration level. To learn more about how to configure integration-level policies for the Gmail integration, read this document.
The steps to configure alert channels for policy-level integration are the same as in the case of integration-level alerts. You can refer to this document for steps.
Learn how to configure risk score and name a Nightfall policy created for ChatGPT.
In this final stage, you assign a name to the policy, verify your configurations, and create the policy.
Enter a name for the policy.
(Optional) Enter a description for the policy.
Choose the Policy risk score. By default the risk score is set to Nightfall Risk Score. You can set it to Custom Risk score, and select one of the risk levels, if required. To learn more about Risk scoring, refer to the #risk-scoring document.
Click Next.
Verify if all the policy configurations are set up as per your requirements.
(Optional) Click back or click on any specific stage to modify any of the policy configurations.
Click Save Changes.
Learn how to configure the detection rules section in Nightfall policies created for ChatGPT.
In this section, you can select the Detection rules for the policy and If not already created, you can create detection rules. To learn more about how to configure detection rules, see Configuring Detection Rules.
You can use the search bar to search for a detection rule by its name.
Once the required detection rules are displayed, you can select the required detection rules by ticking the respective check box. When you select any detection rule, you can view three options.
These three options are related to the display of detection rules.
All Detection Rules: This option displays all the available detection rules, irrespective of the detection rule(s) selected.
Selected Detection Rules: This option displays only those detection rules that you have selected.
Unselected Detection Rules: This option displays only those detection rules that you have not selected.
Select the check box(es) of all the detection rules you wish to include in the policy. The policy evaluates only those detection rules that you have selected here. Once you select all the required detection rules, click Next to move to the next stage.
