Vendor Considerations

With the development of AI, many vendors and political consultants selling technology to campaigns are integrating AI into the products they sell and services they perform. We encourage you to review how your vendors and consultants are using AI on your behalf and if you license technology that includes AI, to consider the following: 

• Who owns the data used or generated by the product? If you are paying for a licensed AI tool, you will want to validate ownership both of any data used by the AI tool or output generated by the product. 
• Will our data be reused? If your organization’s data is used by an AI tool through a technology provider or a consultant, ensure you have clear knowledge if the data is used in any way to train the models or could be used/accessed by another customer.  
• How was the product trained? Transparency is an essential element to building trust in the technology you are using. This includes insights into how the product was developed - models it leverages and data sets that were used to train the models.  
• How is intellectual property handled?: You want assurances that AI outputs aren’t inadvertently or directly including materials subject to third-party copyright or patent protections.
• What is your process of reporting to customers if there is a cyber incident? The vendor should be able to articulate the types of incidents and the speed with which a cyber incident is reported to customers, including how notifications will be made. If the product you are considering is vital and disruptions would have a significant impact, you can ask how the vendor would maintain continuity of service during an incident. 
• Does this product integrate with other tools? There needs to be transparency around how the product interacts with other tools when it comes to your data and the potential for your data to be used to help other customers or train the vendor’s model.