20 January 2024
OpenAI's first account suspension for violating its rules against using ChatGPT APIs for political campaigning occurred on Friday.
According to The Washington Post, AI start-up Delphi used ChatGPT’s conversational software to develop Dean.Bot for political action committee (PAC) We Deserve Better.
The bot mimicked the style of Democratic presidential hopeful Rep. Dean Phillips to converse with potential voters via a website in support of his campaign.
Although the site included disclaimers that the chatbot was an AI tool and not the real Dean Phillips, and the site required user consent to interact with the bot, OpenAI suspended Delphi’s account on Friday.
The suspension was for violating API usage policies against using OpenAI tools for political campaigning or impersonating an individual without their consent.
Delphi took down Dean.Bot shortly after the account was suspended.
Although this may seem a relatively minor misuse of ChatGPT, it is the first time OpenAI have suspended an account for violating its political campaigning rules.
After Argentina’s recent presidential elections saw widespread use of AI-generated content, this is unlikely to be the last time large language models feature in elections during 2024.
As global efforts to regulate AI take shape, its use for political purposes and potential for spreading disinformation can’t be underestimated. Let’s hope global standards emerge to tackle these issues, rather than leaving individual tech companies to set the rules through their internal policies.
As many tech leaders have said, we need to learn the lessons from the rise of social media...