As the 2024 U.S. presidential election approaches, the rapid development of AI technology is garnering increasing attention. Recent analysis by Axios reveals that at least 26 states have either passed or are considering relevant laws to regulate the use of generative AI in elections. This indicates that in future elections, AI will no longer be a simple technological tool but a potential "game-changer" that could influence voting outcomes.
Image source: Image generated by AI, authorized service provider Midjourney
It is well known that generative AI can sometimes be misused, potentially causing voter confusion and suppression. Senator Mark Warner of Virginia pointed out in an email that the risks of AI misuse are significant. He stated, "I believe that companies and platforms developing generative AI do not pay enough attention to its potential for abuse." This has led many to worry about whether the use of AI could affect the fairness of democratic elections.
In fact, there have been multiple instances where generative AI has indeed been used to mislead voters. For example, in New Hampshire, fake calls generated by AI, impersonating Biden's voice, encouraged Democratic voters not to participate in the primary. Last July, a super PAC supporting Florida Governor DeSantis also used AI to mimic Trump's voice in an attack ad. These incidents have made people realize that the misuse of AI is a problem that cannot be ignored.
Although some states have started to take action, such as Texas becoming the first state to ban the creation and distribution of deepfake videos targeting candidates in 2019, at the federal level, regulatory measures regarding AI are still limited. Even President Biden's AI executive order is voluntary and lacks the force of mandatory enforcement.
Currently, 19 states have passed laws specifically targeting AI-generated "deepfake" media. California's law prohibits the dissemination of "substantially deceptive" audio-visual content within 60 days before an election, while Minnesota and Michigan have banned the spread of AI-generated deepfakes within 90 days before an election in 2023. Additionally, New Mexico, Florida, Utah, and others have passed laws this year requiring the disclosure of AI-generated content.
While some states consider exceptions, such as if the content is disclosed or is a satirical or parody advertisement, there are still differences in how states classify and penalize violations. With more states taking action, experts say that to truly address this issue, comprehensive federal action is needed.
** Key Points:**
📊 ** At least 26 states are considering or have passed laws to regulate the use of generative AI in elections.**
📞 **Instances of malicious use of AI technology are common, including fake calls and ads mimicking candidate voices.**
⚖️ ** Currently, federal-level regulatory measures are relatively limited, and state laws vary, necessitating a more comprehensive legal framework.**