According to the OpenAI blog, ChatGPT, Dall-e, and other company AI tools are now prohibited from being used to impersonate candidates or local governments and OpenAI tools can no longer be used for campaigns or lobbying. OpenAI tools are also not allowed to be used to discourage voting or misrepresent the voting process.
In addition to OpenAI now being firmer in its policy on election disinformation, the company also plans to add Coalition for Content Provenance and Authenticity’s (C2PA) digital credentials to images created by Dall-E earlier this year. Currently, Microsoft, Amazon, Adobe and Getty are also working with C2PA to combat misinformation spread through AI image generation.
The digital system would then code the images based on their origin, making it much easier to identify an artificially created image without having to look for odd hands or anything else out of the ordinary.
OpenAI tools will start directing voting questions in the US to CanIVote.org, one of the best places to get information about voting on the Internet.
Source: The Verge
