Grok, distinct from the similarly named AI firm Groq which earlier today disclosed a fundraising of over $600 million, propagated incorrect details regarding Vice President Kamala Harris on X, previously recognized as Twitter.
This incident was highlighted in an open letter drafted by five state secretaries and aimed at Elon Musk, who heads Tesla, SpaceX, and X. The communication criticized Grok, X’s AI-driven chat mechanism, for wrongly intimating that Harris might be disqualified from appearing on certain U.S. presidential ballots for the year 2024.
The initiative, led by Minnesota’s secretary of state Steve Simon and endorsed by peers Al Schmidt (Pennsylvania), Steve Hobbs (Washington), Jocelyn Benson (Michigan), and Maggie Toulouse Oliver (New Mexico), calls on Musk to “promptly make modifications to X’s AI search aide, Grok, to assure voter access to dependable information during this pivotal election cycle.”
On July 21, shortly after President Joe Biden announced his withdrawal from the presidential race, Grok misleadingly began to assert that the ballot submission timelines had elapsed in nine states: Alabama, Indiana, Michigan, Minnesota, New Mexico, Ohio, Pennsylvania, Texas, and Washington.
In reality, these ballot deadlines had not been missed. Nonetheless, Grok’s erroneous dissemination reached a vast audience on X and elsewhere, with corrections only being made on July 31, as noted in the letter.
“Even though Grok is accessible solely to X Premium and Premium+ members and includes a notice encouraging users to double-check information, the incorrect claims regarding ballot deadlines have been repeatedly captured and circulated in several posts,” the state secretaries mentioned.
Musk has faced criticism over X’s approach to political content moderation—and his role in exacerbating the issue.
Evidence points to X having markedly less moderation personnel compared to other platforms, a situation stemming from Musk’s decision to slash approximately 80% of the firm’s engineers focused on trust and safety. Earlier this year, X committed to the creation of a new trust and safety hub in Austin, Texas, yet the recruitment for the center fell significantly short of initial estimates, according to Bloomberg.
Musk has hardly been the model of accuracy enforcement.
Last Friday, the CEO redistributed a manipulated video that artificially reproduced Harris’ voice, falsely depicting her as admitting to being a “diversity hire” and confessing her ignorance in governance matters. The billionaire also asserted “civil war is inevitable” in reaction to riots unfolding across the U.K., ignited by the murder of three girls last week and the ensuing misinformation about the assailant, leading to a stern rebuke from the U.K. Prime Minister Keir Starmer.
Compiled by Techarena.au.
Fanpage: TechArena.au
Watch more about AI – Artificial Intelligence
