With the increasing use of AI-driven features in customer support, data privacy remains a key concern for brands. This is especially critical when children are involved in customer interactions, making compliance with the Children’s Online Privacy Protection Act (COPPA) essential. This includes safeguarding sensitive conversations involving children and preventing unauthorized AI processing of such data.

To meet these regulatory standards, we have introduced the Full Privacy mode (FP). When enabled, FP automatically disables all AI-powered features for flagged interactions, ensuring that AI does not process or generate responses in these cases.

By using Full Privacy mode, brands can enhance child privacy protection and ensure compliance with COPPA regulations.

Features Covered

When Full Privacy mode is turned on, the following AI features will be disabled:

  • AI Sentiment (Sentiment Analysis)
  • AI Summary (Issue Summary)
  • AI Reply (Suggest Response)
  • AI Compose (Smart Compose)

Additionally, issues marked with FP=ON will not be used as training data for AI models or for any other purposes in Helpshift’s database.

Features When FP=ON

Agent Copilot Features (AI Summary, AI Reply, AI Compose, AI Sentiment): When FP=ON, these features will be automatically disabled, and agents will not be able to view or use them.

How to Enable It?

To enable the Full Privacy mode for Issues, follow these steps:

  1. Log in to your Helpshift dashboard.
  2. Go to Issues and open one of the Issues.
  3. Under Metadata, you will see the option to turn on the Full Privacy mode.
     

Note:

  • All data and logs related to FP=ON issues will still be stored in the backend. This allows brands to pull necessary information for compliance tracking and audit purposes while ensuring that FP=ON issues are not used for AI training or any other AI related purposes.
  • If no value is provided for the FP (Full Privacy) flag, the system will assume the issue is not related to a child, and the AI Agent Copilot features will remain enabled. We recommend explicitly setting the FP flag to ON or OFF, based on whether you want these features to be restricted or enabled.