# Understanding the Securing Elections From AI Deception Act

This bill aims to prevent people from using artificial intelligence to interfere with voting and elections. It targets two main problems: using AI to spread false information about voting, and creating fake videos or audio of candidates to mislead voters.

The first part makes it illegal to intentionally use AI to deprive someone of their right to vote. This covers spreading lies about when, where, or how to vote, making false claims about who's eligible to vote, or spreading misinformation about how votes are counted. It also prohibits using AI-generated threats, intimidation, or deception to prevent people from voting, supporting a candidate, or working as election staff. Anyone caught doing this could face up to a year in prison and fines. The law takes effect 60 days after being signed.

The second part requires anyone who creates AI-generated content about elections to include a disclaimer. If you use AI to create fake videos of candidates, fake audio of their voices, or content about voting procedures, you must label it as AI-generated—either with text saying "AI Disclaimer: This output was generated by artificial intelligence" or with an "AI" symbol. The companies that build these AI tools must make it possible for users to add these disclaimers and prevent them from being easily removed. The Federal Trade Commission will write detailed rules about how these disclaimers should work and will enforce the law. This part takes effect 60 days after the FTC issues its regulations.

The bill also requires AI developers and companies using AI to take reasonable safety measures. They must evaluate potential harms before deploying their systems, ensure the AI works as advertised, and consider whether the benefits outweigh the risks. The FTC will enforce these requirements as violations of consumer protection laws.