OpenAI, the artificial intelligence company, recently published a blog post aiming to address concerns that its technology could interfere with elections. As more than a third of the world prepares to head to the polls this year, according to a report by Reuters, the company, backed by Microsoft, has faced scrutiny since unveiling the widely known chatbot “ChatGPT” and the program “DALL-E,” capable of producing realistic deepfake images. The use of artificial intelligence to potentially compromise election integrity has become a worrisome issue.
Among those expressing concerns is OpenAI’s CEO, Sam Altman, who testified before Congress last May, expressing “concern” about the generative AI’s ability to undermine election integrity through “misleading individual interactions.” The company, based in San Francisco, stated that it is collaborating with the National Association of Secretaries of State, a U.S. organization focused on promoting effective democratic processes, such as elections. The United States is set to hold presidential elections this year.
In the blog post, the company mentioned that ChatGPT will direct users to the “Can I Vote Dot Org” website when faced with certain election-related questions. The report also noted that OpenAI is actively working to make the creation of AI-generated images using DALL-E more transparent, planning to include a watermark indicating AI creation. Additionally, they are working on ways to identify content generated by DALL-E even after image modifications.
The company affirmed in the post that its policies prohibit the misuse of its technology, such as creating chatbot robots pretending to be real people or attempting to manipulate individuals’ voting behavior. They also explicitly forbid the use of DALL-E to generate images of real individuals, including political candidates.
Leave a Reply