OpenAI reported rejecting over 250,000 requests on ChatGPT to generate images tied to politicians, enhancing its safety protocols to mitigate deepfake concerns.
In preparation for the elections, OpenAI's strategy aimed to prevent misinformation while guiding users towards accurate voting-related resources like CanIVote.org.
On election day, ChatGPT directed users to trusted news sources for results, avoiding political bias and ensuring factual integrity in its responses.
Despite DALL-E's restrictions, numerous election-related deepfakes circulated, highlighting ongoing challenges in addressing misinformation amid heightened political discourse.
Collection
[
|
...
]