OpenAI suspends developer of politician-impersonating chatbot

23.01.2024

OpenAI has taken decisive action by suspending the developer responsible for Dean.Bot, a ChatGPT-powered chatbot designed to impersonate Democratic presidential candidate Dean Phillips. The bot, developed by cloning startup Delphi, had the intention of supporting Phillips in his political campaign. However, this move directly contradicted OpenAI’s policies, leading to the suspension of the responsible developer.

According to The Washington Post, despite displaying a disclaimer explaining the nature of the chatbot to visitors, Dean.Bot’s purpose was evidently to engage with potential supporters and disseminate the candidate’s message. OpenAI, in response, confirmed the suspension of the developer, emphasizing the violation of their policies. This action aligns with OpenAI’s recent blog post outlining preventive measures against the misuse of its technology, specifically prohibiting “chatbots impersonating candidates” in the lead-up to the 2024 elections.

It’s noteworthy that OpenAI’s policies extend beyond this specific case, explicitly stating that applications for political campaigning and lobbying are not permitted. In a crucial election year, not only in the US but also in other Western democracies like the UK, OpenAI’s firm stance against bots that could be perceived as interference or spreading misinformation is unsurprising.

In response to the inquiry, Delphi initially removed ChatGPT from the bot and continued its operation using alternative open-source tools. However, OpenAI’s intervention ultimately led to the suspension of the chatbot on Friday night. Visitors to the Dean.Bot website are now informed that the chatbot is inaccessible due to “technical difficulties,” accompanied by a message stating, “Apologies, DeanBot is away campaigning right now!”