Primaries, caucuses, conventions, and oh so many campaign ads—the big election year is finally here. This year, though, voters are encountering something that sets this election cycle apart from years past: the rise of generative artificial intelligence.
The field of AI broadly as an academic discipline is decades old, and today many people interact with forms of AI on a daily basis. Non-generative AI is already used in limited ways with heavy human oversight for election administration, such as helping officials answer voter questions, and it’s possible that AI—if deployed responsibly—could continue to help resource-strapped election offices in the future. However, the release of online tools like ChatGPT and DALL-E, which fall into a narrower category known as generative AI, is sparking debate among developers and lawmakers about how to regulate a rapidly evolving branch of AI technology.
As the public’s awareness grows, so do concerns surrounding the use and misuse of generative AI in upcoming elections. A recent poll found that 58% of U.S. adults worry about AI increasing misinformation during the 2024 election cycle, a statistic reflective of justified wariness. After all, it didn’t take sophisticated technology in 2016 and 2020 to manufacture highly compelling but categorically false information about the integrity of U.S. elections.
So, could the high-tech AI version be worse? The answer is “Yes, and.” Yes, the odds of generative AI amplifying misinformation and repackaging it in a more polished form are high, and voters have experience navigating unreliable information ecosystems. Examples of how generative AI could exacerbate existing misinformation tropes include:
- Manipulated images microtargeting specific voting demographics purporting to show voter intimidation at polling locations.
- Fabricated headlines made to look like local news outlets flooding social media showing AI-generated images of illegal poll worker activity.
- AI-generated robocalls impersonating election officials telling voters incorrect information about the voting process to sow confusion
Generative AI isn’t rewriting the misinformation playbook. Still, it is refining it, and barring the deployment of robust detection tools and disclosures for when AI is used in synthetic media, voters must continue to turn to trusted sources for information about elections. Your local elections office is your number one source for information about elections and voting. Instead of asking Bard to find the nearest polling location, consult your local election office’s website or give them a quick call (you can find your local election official’s contact information on the Verifier).
Reputable news outlets have increased their fact-check desks in the wake of the 2020 election, and voters can also consult sites like Snopes and FactCheck.org. Look for information from non-partisan sources you can trust for comprehensive, soup-to-nuts information about voting. We’re proud to help lead the year-round non-partisan Election Protection hotline along with fellow organizations committed to voting rights and sharing accurate information—bookmark all the ways you can contact the 866-Our-Vote hotline.
The dynamics of the 2024 election cycle may feel different, but when it comes to evaluating information, the same rules still apply. Be discerning, seek out non-partisan, data-driven resources, and when in doubt, ask your local election official to help you out.