Government officials caution against relying on AI chatbots for accurate voting information.

• Officials from the U.S. government are urging voters to be cautious about using AI chatbots for election-related queries.
• On Friday, the New York Attorney General Letitia James issued a warning.
• The alert warned that New Yorkers who rely on chatbots risk being misinformed and losing their chance to vote due to inaccurate information.

Government officials are urging against using AI chatbots for voting-related information four days before the presidential election.

The New York Attorney General's office announced on Friday that they had tested various AI-powered chatbots and discovered that they often provided incorrect information about voting when asked sample questions.

In the US, Election Day falls on Tuesday, with Donald Trump, the Republican nominee, and Kamala Harris, the Democratic vice president, in a virtual tie.

Official government sources are more reliable than chatbots for New Yorkers seeking voting information, as the latter may provide inaccurate information that could lead to missing their chance to vote.

The 2020 election cycle is a significant year for political campaigns globally, with over 4 billion people in more than 40 countries being affected by elections. The increasing use of AI-generated content has raised concerns about the spread of election-related misinformation.

According to Clarity, a machine learning firm, the number of deepfakes has increased by 900% annually. Some of these videos were created or funded by Russians with the intention of disrupting the U.S. elections, as stated by U.S. intelligence officials.

There are concerns among lawmakers about the spread of misinformation resulting from the emergence of generative AI, which became widely available in late 2022 with the launch of OpenAI's ChatGPT. Despite being a relatively new technology, large language models are known to frequently produce inaccurate and unreliable information.

"Alexandra Reeve Givens, CEO of the Center for Democracy & Technology, advised voters not to rely on AI chatbots for voting information due to concerns about accuracy and completeness. Numerous studies have demonstrated instances of AI chatbots providing incorrect information about polling locations, voting accessibility, and acceptable voting methods."

A July study by the Center for Democracy & Technology found that over one-third of AI chatbot responses to 77 election-related queries contained incorrect information. The study tested chatbots from Mistral, , OpenAI, and Anthropic.

ChatGPT will provide a message encouraging users to check news sources like the Associated Press and Reuters or their state or local election board for the most complete and up-to-date election results starting on November 5th.

OpenAI disrupted over 20 operations and deceptive networks worldwide that attempted to use its models, according to a 54-page report published last month. The threats included AI-generated website articles and social media posts by fake accounts, though none of the election-related operations were able to attract "viral engagement."

Since November 1st, Voting Rights Lab has monitored 129 bills across 43 state legislatures that aim to control the possibility of AI generating election misinformation.

WATCH: More than a quarter of new code is now AI-generated