Voters face uneven AI deepfake protections

As Americans prepare for the upcoming midterms, the potential impact of AI-generated deepfakes on their voting experience is becoming increasingly complex. This is due to a lack of uniform regulations across states, resulting in a patchwork of laws that could lead to vastly different outcomes depending on where an individual resides.

Twenty-nine states have enacted laws related to AI-generated election deepfakes, while California and Hawaii have had their respective laws struck down by courts. Minnesota and Texas prohibit the use of political deepfakes within a certain time frame before an election, while Maryland's ban is indefinite. The remaining states mandate disclosures whenever AI is utilized in political advertisements.

Some jurisdictions, such as Colorado and Utah, require more granular disclosures including the creator of the deepfake, the date of creation, and the editing process.

The conflict arises from the First Amendment challenges faced by California and Hawaii's AI deepfakes laws. Both states are now pursuing alternative strategies, with California implementing the AI Transparency Act on August 3, which requires disclosures of AI-generated images. Hawaii recently passed a law that prohibits the creation of deepfakes in political ads without the subject's consent.

At the federal level, lawmakers have yet to establish a standardized approach to dealing with AI deepfakes in elections. While Congress' initial AI deepfakes law, the Take it Down Act, focused on non-consensual intimate imagery, it remains to be seen how effectively this law will be implemented. Further legislation aimed at election deepfakes is expected if Democrats regain control of the House next year.

The disparity in state regulations means that voters could encounter a range of experiences regarding AI deepfakes leading up to Election Day. This uneven landscape presents a significant challenge for ensuring a fair and transparent electoral process.