OpenAI removed VIC, an AI chatbot that was running for mayor in Wyoming, from its public-facing service.
This move was done on the ground that the AI bot violated the policies of the company regarding political campaigning.
OpenAI Removes AI Running For Mayor
Victor Miller, a 42-year-old man from Cheyenne, Wyoming, recently filed an application for mayor. However, this application was not his. Rather, he listed a customized AI chatbot's name to be included in the ballot.
Miller claimed that the AI bot, dubbed Virtual Integrated Citizen (VIC), would capitalize on OpenAI's technology to come up with political decisions and help in governing the city. VIC has not been tied to any political party.
He shared that as AI has helped him personally, he thinks that the technology could also help the town, adding that he would like to see it happen.
However, OpenAI has recently confirmed that it has taken action against the AI chatbot, as VIC has gone against the firm's political campaigning guidelines.
OpenAI noted in a blog post from January that it is now allowing political groups or politicians to make use of their technology to come up with campaign materials. OpenAI also says that a fully AI politician is covered by such a bill.
While this is a loss for VIC and Miller, this is a win for Chuck Gray, Wyoming's Secretary of State, who has claimed that AI does not meet the qualifications for electoral bid as it lacks being a human within a physical body.
Gray explained that the law of Wyoming is clear: in order to run for office, the candidate should be a qualified elector. This necessitates that the candidate should be a real human being.
While VIC is no longer publicly accessible, Miller can still access the bot. He also plays to invite residents of Cheyenne to interact with the AI bot during a meet-up in a local library.
VIC still has another major challenge to face. Local authorities are still looking into whether the lack of physical form makes VIC disqualified. If they conclude that VIC's electoral run goes against the law, Miller may have other more grave matters to face.
OpenAI Takes Action Against AI Model For Parliament Campaign
On top of this issue with VIC, OpenAI has also taken action against another UK candidate who made use of AI models for Parliament campaigning.
Steve Endacott, the chairman of Neural Voice, an AI company, caters to questions from voters through the chatbot "AI steve" on his website.
Endacott has been running independently and has not commented on the matter.
On his website, a ChatGPT chatbot was offered wherein voters could share opinions and help with policy creation. If voters asked a question regarding policies and the AI could not answer, the bot would perform a search and come up with a policy suggestion.
Though Endacott's website is still functional, the bot is no longer ChatGPT-powered.
AI in Politics
Though AI bots are becoming smarter, experts think that technology should never replace human judgment when it comes to running any area of the government.
Jen Golbeck, a professor from the University of Maryland's College of Information Sciences, explained that when it comes to AI at present and how it would fare in the future, this technology must never be used for coming up with automated decisions.
AI was designed to support decisions, as it could offer data to help humans with decision-making. However, AI is not designed to make decisions independently.
AI political candidate emergence comes in the middle of rising concerns regarding how misinformation spread may affect elections.
However, Golbeck said that AI may have a place in politics, such as when answering constituents' forms or directing some problem-solving.
Golbeck noted that, while a chatbot can be trained with knowledge in the office, humans should always be the decision-makers.
RELATED ARTICLE: AI in Court Hearings? Judges in England, Wales Can Now Use ChatGPT For Legal Rulings
Check out more news and information on Tech & Innovation in Science Times.