The United States is in the midst of a highly competitive presidential election. In this regard, misinformation spreads easily due to access to artificial intelligence. Minnesota Secretary of State Steve Simon has reported on an incident that shows this problem: his office discovered election misinformation coming from X’s Grok chatbot.
Grok AI chatbot disinforms about US elections
The Minneapolis Star Tribune reported that Grok, X’s premium AI chatbot service, was spreading incorrect information about presidential voting deadlines in several states, including Minnesota.
When asked about the deadlines, Grok wrongly claimed that the states of Alabama, Indiana, Michigan, Minnesota, New Mexico, Ohio, Pennsylvania, Texas, and Washington had already passed their presidential votes. This false information came from an X post made by conservative user @EvanAKilgore, which had no correction or warning about the misinformation.
Things get worse when Grok, in his “Funny Mode,” repeats the incorrect information and source with inappropriate humorous comments.
For example, Grok joked, “So if you’re thinking about running for president in one of these states, you might want to check if there’s a write-in voting option. Or, you know, travel back in time to before the deadlines passed.”
In another response, the chatbot implied that President Joe Biden’s name couldn’t be replaced on these states’ ballots because the deadlines had already passed. However, these claims are completely false.
The reality…
President Biden was the presumptive nominee until he announced he would not seek a second term. In reality, the Democratic Party will choose its nominee via virtual roll call voting before its convention, and states don’t begin printing ballots until after both conventions. Additionally, in Minnesota, the deadline for voting is not until August 26, flatly debunking Grok’s claims.
Fact checks by reputable media outlets, such as Politifact, have refuted these claims. Politifact rated a similar claim, with the same list of nine states, as “False.”
Steve Simon also commented on the difficulties encountered in communicating the misinformation to X. According to Simon, his team’s attempts to contact a media relations representative for X were unsuccessful, and the response they received was unsatisfactory and concerning. “They received what I can only verbalize as the equivalent of a shrug,” Simon told the Star Tribune.
When the Minneapolis Star Tribune attempted to get comment from X, they only received a short, evasive response that read, “Busy now, check back later.”
Actions to take
To avoid misinformation on social media, chatbots, the internet and other media, you need to develop a critical and conscious approach to the information you consume.
First, verify the source of information before trusting or sharing it. This involves checking the credibility of the author, the date of publication and whether the information has been corroborated by reliable sources. Fact-checking platforms, such as Politifact or Snopes, can be useful in confirming the veracity of the data. Also, be skeptical of sensationalist claims or claims that seem too good or bad to be true, as these are common characteristics of misinformation.
Second, diversify information sources. Relying exclusively on a single outlet or platform can limit perspective and increase susceptibility to misinformation. If you consume news and data from a variety of sources, including international and local, you can get a more balanced and complete view of events. Use news aggregation tools and follow journalists and experts from different fields to gain a broader and more nuanced understanding of the issues.
Finally, foster critical thinking and media literacy to combat misinformation. This includes learning to distinguish between facts and opinions, identifying biases, and understanding how media narratives are constructed.