ChatGPT maker OpenAI rolls out plans to fight election disinformation

ChatGPT maker OpenAI said it is gearing up to thwart disinformation as dozens of nations head to polls this year.

OpenAI said it wants to ensure its technology is not used in a way that could undermine the election process, including deploying measures to prevent abuses such as deepfakes and chatbots impersonating candidates.

The surge in ChatGPT’s success sparked off a global AI revolution, which has triggered warnings of the usage and spread of disinformation tactics to sway voters.

“As we prepare for elections across the world’s largest democracies, our approach is to continue our platform safety work by elevating accurate voting information, enforcing measured policies, and improving transparency,” the company said in a blog post.

OpenAI said it wants to ensure that its AI systems are built, deployed, and used safely.

Misuse that undermines democratic participation, such as spreading false voting information, is also not allowed, with new GPT versions including features for users to report violations, it said.

“We have a cross-functional effort dedicated to election work, bringing together expertise from our safety systems, threat intelligence, legal, engineering, and policy teams to quickly investigate and address potential abuse,” it said in the blog post.

Prior to launching new systems, OpenAI said it conducts thorough testing, gathers feedback, and implements safety measures to minimize potential harm.

OpenAI said it is continuously updating its usage policies for ChatGPT to address potential misuse.

“Protecting the integrity of elections requires collaboration from every corner of the democratic process, and we want to make sure our technology is not used in a way that could undermine this process,” it said.

The ChatGPT maker said it has developed tools to enhance factual accuracy and reduce bias, with DALL·E having safeguards against generating images of real people, including candidates.

OpenAI said it is working on improving transparency around AI-generated content and implementing digital credentials to track the provenance of images. ChatGPT will soon integrate real-time news with clear attribution to aid users in information assessment, it added.

In the US, OpenAI said it collaborates with the National Association of Secretaries of State to guide users to reliable voting information on CanIVote.org. These efforts will inform strategies in other regions, it added.