AI Use Policy Framework

When thinking about an acceptable use policy for your campaign consider the following elements:

• Principles, Values, Purpose, and Scope
• Benefits and Risks
• Dos and Don'ts of AI use
• Implementing Your Policy

Below are some questions within these areas for consideration as you develop and shape your policy. Not every question is intended to result in a section of policy. Some, for example, values questions, might be used for framing how you communicate the policy to your team.

• Principles, Values, Purpose, and Scope: AI is a new technology. In the minds of your team and the community, there are likely to be varying levels of understanding about its use and impact. As a candidate, it is important to establish the principles, values, purpose, and scope of your policy to set the tone of how AI is implemented and align the use of AI with your beliefs within your campaign.
  • What are the ethical and moral standards that you adhere to when using AI? 
  • How do you ensure that your AI use is fair, transparent, explainable, and accountable? 
  • How is human oversight, judgment, and feedback incorporated into your AI use? 
  • How do you communicate and explain your AI use and outputs to your stakeholders and/or the public?

• Benefits and Risks: Identify the potential risks and benefits of your campaign’s generative AI use. How you frame this relationship in conjunction with other elements of your policy, such as values, will directly inform the acceptable AI practices that you outline in the policy. 
  • How do you articulate, address, and mitigate the risks and harms of your AI use (such as data privacy, security, accuracy, and unintended consequences)?
  • How do you frame the possible positive and negative impacts of your AI use? 
  • How do you measure and evaluate the effectiveness and performance of AI as it relates to outcomes you have attempted to achieve when implemented?

• Dos & Don’ts of AI Use: The core of an acceptable use policy is the clear roadmap of practices that are a logical extension of your principles, values, purpose, and scope, and align with the benefits and risks you have articulated. This can include: 
  • What are the specific rules and guidelines that you follow when using AI? 
  • What AI tools and applications are you comfortable using to ensure compliance with your organization’s data security policies? 
  • When can your organization use public generative AI tools and are there specific restrictions when using these public tools?  
  • Are there specific disclosure and labelling practices you want staff using for AI-generated content (internally and externally)?  
  • What is the approval process for the dissemination of AI generated content?
  • How do you monitor and audit the quality and accuracy of your AI outputs?

• Implementing Your Policy: Thought should be given to how the policy will be implemented so your team moves forward in achieving the policy’s goals. 

• How will you introduce your policy to your team and during onboarding of new staff?
• How will you monitor ongoing compliance with policy?
• Who is the person that people should turn to with questions or concerns?
• How do you select and vet any new AI tools and vendors moving forward?