Federal Election Commission - Artificial Intelligence and Campaign Ads
Federal Election Commission
[From the NTIA - Aug 11, 2023]
REG 2023-02 (Artificial Intelligence in Campaign Ads) Draft Notification of AvailabilityThe Commission approved a Draft Notification of Availability in response to a Petition for Rulemaking filed by Public Citizen. The Petition asks the Commission to amend its regulation on the fraudulent misrepresentation of campaign authority to make it clear that the related statutory prohibition applies to deliberately deceptive Artificial Intelligence (AI) campaign advertisements. The Commission seeks written comment on the Petition. The Notification will be published in the Federal Register at a future date along with the deadline for comments.
News
- AP, FEC moves toward potentially regulating AI deepfakes in campaign ads, Aug. 10, 2023
- CNN, Federal regulators inch closer to regulating AI in political ads, Aug. 10, 2023
- The Hill, FEC to consider new rules for AI in campaigns, Aug. 10, 2023
- Public Citizen, Petition To Regulate “Deepfake” Election Ads Will Proceed to Public Comment Period, Aug. 10, 2023
About the Federal Election Commission
The FEC was created to promote confidence and participation in the democratic process. The mission of the FEC is to protect the integrity of the federal campaign finance process by providing transparency and fairly enforcing and administering federal campaign finance laws.
Draft Notification of Availability
Draft Notification of Petition for Rulemaking
CAIDP Comments to FEC on AI and Campaign Ads (Oct. 13, 2023)
On October 13, 2023, the Center for AI and Digital Policy submitted detailed comments to the Federal Election Commission regarding the use of AI for campaign advertising. Here are excerpts:
CAIDP supports the petition for rulemaking submitted by Public Citizen. The Commission should amend its regulation on fraudulent misrepresentation of campaign authority to make clear that the related statutory prohibition applies to deliberately deceptive Artificial Intelligence campaign advertisements and public communications.
In our comment below we underscore the unique risks of generative AI and machine learning systems and how these unique risks set these techniques apart from other communications and social media technologies that may be used for campaign ads.
We recommend (1) the FEC issue a policy statement pertaining to “fraudulent misrepresentation” arising out of the use of generative AI systems and (2) the FEC assign legal liability to actors who use generative AI in “public communications” and campaign advertisements without attribution.
Unique Risks of Generative AI
- Generative AI has a high propensity for producing indistinguishable falsehoods
- Generative AI can mimic and manipulate human behavior
- Generative AI can amplify biases and political microtargeting.
- Generative AI can be deployed in many levels of the election process
Transparency is a fundamental obligation in Established AI Governance Frameworks
- OECD AI Principles (endorsed by the US)
- UNESCO AI Recommendation (US rejoined UNESCO)
- Universal Guidelines for AI (widely endorsed by AI experts and legal scholars)
Complete CAIDP Statement to the FEC on AI and Campaign Ads [download]
Public Citizen Petition
Public Citizen launched the campaign that resulted in the decision of the FEC to begin the rulemaking on AI and Campaigns Ads.
- Public Citizen, Petition for Rulemaking to Clarify that the Law Against ‘Fraudulent Misrepresentation’ Applies to Deceptive AI Campaign Ads, May 16, 2023
- Public Citizen, FEC to Consider Rulemaking on Political Deepfakes in Response to Public Citizen Petition, June 20, 2023
- Public Citizen, Public Citizen Submits Petition to Federal Election Commission Calling for Rulemaking on AI, July 13, 2023
- Public Citizen, Petition To Regulate “Deepfake” Election Ads Will Proceed to Public Comment Period, Aug. 10, 2023
Letter from Congress to FEC in Support of Public Citizen Petition
