ChatGPT denied 250,000 election deepfake requests.

OpenAI’s DALL-E image generator and ChatGPT have put in place a system that deters deepfake creation in the US elections. For the sake of preserving the fairness of all elections and fighting all forms of speculative propaganda, OpenAI, however, forbade the generation of images of such politicians as President Biden, President-elect Trump, Vice President Harris, Vice President-elect Vance, and Governor Walz, turning down more than 250,000 such requests.

In fact, the company’s own internal policies concerning the safety of their users and the content accessible to them proved to be very strict as regards the requests for images of supporting living persons, especially politicians.

OpenAI’s preparations to safeguard the United States presidential elections from being rigged began early before the date of the election. The rate of disinformation on the net increased from the beginning of the year, and the company has been making advance work to address the concerns.

An order to discourage possible disruption of the voter process principles, including misinformation, was another step in this direction. Such an initiative was designed to avoid the potential for confusion resulting from conflicting facts and instead assist people in accessing appropriate information from credible channels, particularly with regard to the election process.

Within the last month before the election day, about one million inquiries in which OpenAI’s ChatGPT included a link to CanIVote.org were handled. This effort was significant in terms of targeting people with information related to their rights with respect to casting votes, as well as their other voting procedures.

On the election day and the following day, ChatGPT offered 2 million responses, which also encouraged users to check out other sources for news, such as the Associated Press, Reuters, and other news platforms. Apart from that, OpenAI tried to be objective by making sure in all possible ways that there were no political sentiments in any of the responses given by the ChatGPT system.

Thank you for reading this post, don't forget to follow my whatsapp channel