OpenAI shares strategy to minimise AI intervention in 2024 elections

Al in 2024 elections: OpenAI has unveiled a plan to ensure the integrity of global elections as the world gears up for crucial polls in 2024. The strategy encompasses several key initiatives aimed at curbing the potential abuse of OpenAI’s technology during elections. One focal point is the proactive prevention of abuse, with a specific focus on countering misleading content such as "deepfakes," scaled influence operations, and chatbots impersonating candidates. The company employs red teaming, user engagement, and safety mitigations to anticipate and address risks associated with their AI systems. 

The company iterates its continuous efforts to refine usage policies for their AI models, including ChatGPT and the API. 

Notably, the prohibition of applications for political campaigning and lobbying aims to prevent any undue influence. The company places emphasis on ensuring that users can trust the interactions they have, disallowing the creation of chatbots pretending to be real individuals or institutions.

In the official blog shared by OpenAi, the company shared its attempts to improve image provenance including implementing the Coalition for Content Provenance and Authenticity’s digital credentials for images generated by DALL·E 3. The company is experimenting with a provenance classifier, a tool designed to detect manipulated images, with plans to involve journalists, platforms, and researchers for feedback.

Furthermore, OpenAI shared that will improve access to authoritative voting information. In collaboration with the National Association of Secretaries of State (NASS) in the United States, ChatGPT will direct users to CanIVote.org, a trusted source for US voting information, when asked specific procedural election-related questions.