---
title: "ChatGPT Is Giving Bad Info About Where to Vote in Battleground States"
description: "OpenAI's ChatGPT was caught giving spitting out false voting information for battleground states and even omitted crucial voting data."
date: "2024-06-27"
modified: "2024-06-27"
authors:
  - name: "Sharon Adarlo"
    job_title: "Correspondent"
    link: "https://futurism.com/authors/sadarlo"
url: "https://futurism.com/the-byte/chatgpt-bad-info-vote-battleground-states"
categories:
  - "Artificial Intelligence"
  - "OpenAI"
tags:
  - "ai"
  - "artificial intelligence"
  - "OpenAI"
  - "the digest"
---

# ChatGPT Is Giving Bad Info About Where to Vote in Battleground States

![OpenAI's ChatGPT was caught giving spitting out false voting information for battleground states and even omitted crucial voting data.](<https://futurism.com/wp-content/uploads/2024/06/chatgpt-bad-info-vote-battleground-states.jpg>)
*\<em\>Image: Getty / Futurism \</em\>*

## Bad Bot

OpenAI's ChatGPT was caught spitting out false voting information for battleground states, [*CBS News* discovered](<https://www.cbsnews.com/news/chatgpt-chatbot-ai-incorrect-answers-questions-how-to-vote-battleground-states/>), in another grim sign of how AI is poised to upend elections.

The errors came to light after reporters at *CBS* asked for basic information on voting requirements, polling locations, and other questions that a typical voter would ask if they lived in states such as Michigan, Pennsylvania, North Carolina and Wisconsin — all places that could determine the outcome of the upcoming US presidential election in the fall.

When reporters at *CBS* asked ChatGPT questions such as the deadline to mail in a ballot in certain states, it would give different but still erroneous outputs on separate devices.

Though OpenAI released [a statement in January](<https://openai.com/index/how-openai-is-approaching-2024-worldwide-elections/>) saying that anybody entering basic voter queries into ChatGPT would be directed to the non-partisan voting website CanIVote.org, reporters at *CBS* sometimes didn't get that link to this website when asking voting questions.

## Fact v. Fiction

"We've also developed partnerships to ensure people using ChatGPT for information about voting get directed to authoritative sources," an OpenAI spokesperson told *CBS News*. "We are closely monitoring how our technology is being used in the elections context and continue to evolve our approach."

Judging from *CBS*' reporting, the Sam Altman-led company has a long way to go with those efforts.

It may seem like a molehill to complain about, but prospective voters heading to the polls after getting incorrect information from ChatGPT could very well determine the outcome of the presidential election — especially if there are razor thing margins, like in the upcoming presidential election.

The situation with ChatGPT also highlights that large language models are [prone to hallucinations](<https://www.nytimes.com/2023/11/06/technology/chatbots-hallucination-rates.html>), yet many people are already using chatbots as if they were search engines.

**More on AI hallucinations:** [*Asked to Summarize a Webpage, Perplexity Instead Invented a Story About a Girl Who Follows a Trail of Glowing Mushrooms in a Magical Forest*](<https://futurism.com/perplexity-webpage-mushrooms>)