ChatGPT rejected more than 250,000 AI image requests in the month leading up to Election Day, according to a blog post on Friday. OpenAI.
The user attempted to generate photos of President-elect Trump, Vice President Harris, Vice President-elect Vance, President Biden, and Governor Tim Walz (D-Minnesota).
“We have applied security measures to ChatGPT and denied requests to generate images of real people, including politicians,” the company wrote. “These guardrails are especially important in the election context and are a critical part of our broader efforts to prevent our tools from being used for deceptive or harmful purposes.”
Instead, ChapterGPT submitted election-related questions to CanIVote.org through a partnership with the National Association of Secretaries of State (NASS) to remain politically neutral.
The platform directs inquiries about election results to news organizations such as the Associated Press and Reuters. In recent weeks, OpenAI has had to ban a foreign influence operation known as “Storm-2035” that pushes out Iranian-influenced political content.
The company said it will continue to monitor ChatGPT to ensure responses are accurate and ethical. Earlier this year, the company praised the Biden administration's efforts to outline policy on national security and AI technology.
