Social Platform X Modifies AI Chatbot Following Election Misinformation Warnings

Social media platform X has modified its AI chatbot following concerns from five secretaries of state about election misinformation. Officials from Michigan, Minnesota, New Mexico, Pennsylvania, and Washington alerted Elon Musk that the chatbot, Grok, was providing incorrect details about state ballot deadlines after President Joe Biden exited the 2024 presidential race.

Election Misinformation Concerns

The secretaries of state urged that the chatbot direct users to CanIVote.org for election-related queries. This site is managed by the National Association of Secretaries of State. Now, before giving responses to election questions, Grok advises users to visit Vote.gov for accurate information about the 2024 US Elections.

In a joint statement, the secretaries of state said both websites are "trustworthy resources that can connect voters with their local election officials." They expressed appreciation for X's efforts to enhance their platform and hoped for continued improvements to ensure users access accurate information during this critical election year.

Grok's Reach and Impact

Grok is accessible only to subscribers of X's premium versions. However, the secretaries noted that misinformation from Grok has been shared widely across various social media platforms, reaching millions. They highlighted that Grok continued spreading false information for ten days before corrections were made.

The platform did not respond to requests for comments on these issues. The recent change directing users to official voting websites does not address Grok's potential to generate misleading AI-created images related to elections. Users have exploited this tool to flood the platform with fake images of candidates like Vice President Kamala Harris and former President Donald Trump.

AI Chatbot Limitations

Grok was introduced last year for X premium and premium plus subscribers. Musk described it as a "rebellious" AI chatbot capable of answering "spicy questions" rejected by other AI systems. Social media platforms have faced increasing scrutiny over their role in spreading misinformation, particularly concerning elections.

The letter from the secretaries also cautioned that inaccuracies are common in AI products, especially chatbots like Grok based on large language models. Since Musk acquired Twitter in 2022 and rebranded it as X, watchdog groups have raised alarms about increased hate speech and misinformation on the platform, alongside staff reductions in content moderation.

Experts argue these developments represent a step back from progress made by social media platforms in combating political disinformation since the 2016 U.S. presidential election. They warn this could lead to a worsening landscape of misinformation ahead of this year's November elections.

"We appreciate X's action to improve their platform and hope they continue to make improvements that will ensure their users have access to accurate information from trusted sources in this critical election year," said the secretaries of state.