Robert F. Kennedy Jr.’s AI Chatbot, Powered by Microsoft, Returns Online

Microsoft’s Azure OpenAI Service Enables Kennedy Campaign to Bypass OpenAI’s Ban on Political Chatbots

In a surprising turn of events, Robert F. Kennedy Jr.’s presidential campaign has managed to create a chatbot despite OpenAI’s strict prohibition on using their technology for political purposes. The campaign leveraged a loophole by utilizing Microsoft’s Azure OpenAI Service through a third-party provider, LiveChatAI. This approach allowed them to access OpenAI models while circumventing the ban, as neither Microsoft nor LiveChatAI explicitly forbid campaigns from using their products.

Chatbot Raises Concerns About Disinformation and Computational Propaganda

The Kennedy campaign’s chatbot, trained on material from their website, has been found to relay information related to Kennedy’s amplification of conspiracy theories. When asked about the CIA’s involvement in the assassination of former president John F. Kennedy, the chatbot responded that based on the provided context, Robert F. Kennedy Jr. believes in the conspiracy theory. This incident highlights the potential for AI to contribute to the spread of disinformation and computational propaganda, as noted by Sam Woolley, the director of propaganda research at the University of Texas at Austin’s Center for Media Engagement:

“This is exactly the type of use of AI that could lead to the proliferation of disinformation and computational propaganda.”

LiveChatAI’s Role in Powering the Kennedy Campaign’s Chatbot

While the Kennedy chatbot page does not disclose the underlying model powering it, the site’s source code connects the bot to LiveChatAI, a company that advertises its ability to provide GPT-4 and GPT-3.5-powered customer support chatbots to businesses. LiveChatAI cofounder Emre Elbeyoglu stated that the platform utilizes various technologies, including Llama and Mistral, in addition to GPT-3.5 and GPT-4. However, due to client confidentiality, Elbeyoglu could not confirm or deny the specifics of the Kennedy campaign’s usage.

OpenAI’s Stance and Microsoft’s Involvement

OpenAI spokesperson Niko Felix indicated that the company had no direct evidence of the Kennedy campaign chatbot building on its services but suggested that LiveChatAI might be using one of its models through Microsoft’s services. Microsoft has been investing in OpenAI since 2019 and has incorporated the company’s language models into various products, such as the Bing search engine.

Chatbot’s Temporary Disappearance and Reappearance

On Sunday afternoon, the chatbot service became unavailable on the Kennedy campaign site, displaying a red exclamation point icon and the message “Chatbot not found.” However, by Monday afternoon, the chatbot had reappeared. WIRED reached out to Microsoft, OpenAI, LiveChatAI, and the Kennedy campaign for comment on the chatbot’s apparent removal and reappearance but did not receive an immediate response.

The Controversy Surrounding Chatbots in Political Contexts

Given the tendency of chatbots to hallucinate and hiccup, their use in political contexts has been controversial. Currently, OpenAI is the only major large language model provider to explicitly prohibit its use in campaigning. Other companies, such as Meta, Microsoft, Google, and Mistral, have terms of service that do not directly address politics. The fact that a campaign can access GPT-3.5 and GPT-4 through a third party without consequence raises questions about the effectiveness of OpenAI’s policies and the limitations on the use of AI in political campaigns.

“OpenAI can say that it doesn’t allow for electoral use of its tools or campaigning use of its tools on one hand. But on the other hand, it’s also making these tools fairly freely available. Given the distributed nature of this technology one has to wonder how OpenAI will actually enforce its own policies.”

Updated: 3/4/2024, 6:38 pm EST: The RFK Jr. campaign’s chatbot reappeared Monday afternoon.