This article is Part 1 of a Democracy Works series about the role of AI in the 2026 midterm elections. Today’s post explores how false and misleading information about elections, including AI-generated content, affects voters and shapes our democracy.
The internet has reshaped every aspect of American democracy. Social media and AI platforms have equipped voters with new options for obtaining and sharing election information. Americans are more interconnected than ever before.
The same technologies that make voter participation easier can also exacerbate voter confusion. In recent years, false and misleading information about elections has spread widely in the US and around the world. AI platforms designed to make information more accessible have provided biased or incomplete responses about voting and elections.
These contradictions aren’t slowing down. The 2026 midterm elections will be the first US federal election where AI systems are deeply integrated into how voters access information, from chatbots to AI features embedded across search engines, social media, and other platforms. The stakes for voters—and our democracy—are rising.
The need for information integrity in AI is an election integrity issue. This three-part content series will grapple with the relationship between AI and elections—and what it means for voters. We’ll diagnose the challenges, survey the current landscape, and identify opportunities to build a stronger elections information infrastructure.
Why Does Accurate Election Information Matter?
The voter journey does not begin and end on Election Day. Accurate election information is critical at every stage of the electoral process, from understanding dates and deadlines to finding where to vote and interpreting election results. With more than 10,000 election jurisdictions in the US, American election administration is decentralized and complex. Because participation can look different depending on where someone lives, voters need trusted information to navigate localized voting processes with confidence.
False and misleading information about elections increases voter confusion and erodes public trust in the democratic process. Americans have little confidence that AI will improve the situation. A March 2026 survey of US adults found that 85% of respondents believe that AI-generated content will spread false information about this year’s midterm elections, while 26% identified misleading information as the most significant threat to safe and secure elections.
Democracy depends on the robust participation of eligible voters and the legitimacy of outcomes. To navigate US elections today, voters must overcome an information space polluted by incorrect election information and false claims about voting. With so much noise, election officials, tech platforms, and civic technologists must confront a key challenge: ensuring voters have access to reliable voting information, wherever they search for it.
A Dual Challenge of Mis- and Disinformation
Both misinformation and disinformation distort the election information space. While there are important differences between them, both are harmful to democracy and can be worsened by insufficient AI guardrails or inadequate information infrastructure.
Misinformation is incorrect information shared without any intent to knowingly deceive others. During an election, sharing false information about voting or candidates—even when unintentional—can confuse voters and prevent them from participating. Ahead of a May 2026 election in Scotland, researchers found that AI chatbots incorrectly answered 34% of election-related questions. The chatbots confused candidates, misstated ID requirements, and provided the wrong election date, often while supplying outdated or broken source links. When well-intentioned AI assistants lack access to authoritative data, they are ill-equipped to explain local election nuances and can confidently misinform voters.
On the other hand, disinformation is false information that is intended to mislead others. In an electoral context, bad actors can use generative AI tools to create deepfakes or other manipulated images about candidates and voting processes. Particularly when deepfakes are used to impersonate candidates, this content aims to change voters’ perceptions and influence election outcomes. As AI-generated content becomes more sophisticated—and more difficult to detect—it is critical for AI tools to be responsibly designed with effective safeguards against misuse.
Preparing for 2026 Elections
This year, AI will impact elections in democracies around the world. As AI tools become more deeply integrated within society, approaches to AI and elections must balance free speech protections with genuine safeguards for information integrity. AI can be a strong participatory tool for democracy, but not at the expense of accuracy, trustworthiness, and accountability. Public confidence in our elections depends on it.
The next post in this series will dive into these complexities by exploring opportunities and challenges for AI during the 2026 US elections. When American voters turn to AI tools for election information this year, what awaits them? Sign up for our newsletter to receive an email with the next post in your inbox.
.png)
