OpenAI, the organization behind the widely used text generator ChatGPT, has announced proactive measures to counter the potential misuse of its AI tools in the context of global elections. With upcoming elections in major countries such as the United States, India, and Britain, OpenAI emphasized its commitment to ensuring that its technology is not exploited to undermine democratic processes.
One significant decision is the restriction on the use of OpenAI’s technology, including ChatGPT and the image generator DALL-E 3, for political campaigns. This move reflects concerns about the potential for AI-driven disinformation to flood the internet and influence voters.
The World Economic Forum recently highlighted AI-driven disinformation and misinformation as major short-term global risks that could pose a threat to newly elected governments in major economies. The fear is that AI-generated content could be used to manipulate public opinion, especially when users struggle to distinguish between genuine and manipulated content.
OpenAI acknowledged the challenge of understanding how effective its tools might be for personalized persuasion and, until more is known, has decided not to allow the creation of applications for political campaigning and lobbying.
In response to the growing threat of AI-driven disinformation, OpenAI is actively working on tools that will enhance the accountability of AI-generated content. Specifically, it aims to attach reliable attribution to text generated by ChatGPT. Additionally, users will be provided with the capability to determine whether an image was created using DALL-E 3.
OpenAI plans to implement the Coalition for Content Provenance and Authenticity’s (C2PA) digital credentials, which encode details about content provenance using cryptography. The C2PA, consisting of industry members like Microsoft, Sony, Adobe, Nikon, and Canon, seeks to improve methods for identifying and tracing digital content.
ChatGPT will also direct users to authoritative websites when asked procedural questions about U.S. elections, ensuring accurate and reliable information dissemination. OpenAI is keen on applying the lessons learned from this initiative to its approach in other countries and regions.
This announcement follows similar efforts by major tech companies, including Google and Meta, to limit election interference and combat the spread of misinformation, especially through AI-driven tools. The move reflects a broader industry recognition of the need to address the risks associated with AI in the context of elections and public information dissemination.
