Ahead of the upcoming general elections, leading artificial intelligence research organization, OpenAI, has announced that it will prohibit the use of its AI technology for political campaigning and will continue its efforts to combat misleading ‘deepfakes’ and chatbots impersonating candidates. The firm, led by Sam Altman, outlined a series of policy changes to ensure that its generative AI-based technologies, including ChatGPT and Dall-e, do not undermine the democratic process during future elections.
In a blog post, OpenAI stated, As we prepare for elections in 2024 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 organization is committed to preventing any abuse of its AI systems that could compromise the integrity of the electoral process, emphasizing the importance of collaboration from all stakeholders to protect the democratic process.
OpenAI highlighted several measures it has implemented to prevent misuse of its technology. The organization conducts rigorous testing and gathers feedback from users and external partners to ensure the safety of its AI systems. It also incorporates safety mitigations to reduce potential harm and improve factual accuracy, reduce bias, and decline certain requests. For instance, Dall-e, their image-generation system, includes guardrails to decline requests for the creation of images depicting real individuals, including candidates.
The organization acknowledged that it is still in the process of understanding the potential influence of its tools for personalized persuasion. Consequently, OpenAI does not currently allow the development of AI applications for political campaigning and lobbying until a more comprehensive understanding is achieved. OpenAI also prohibits the creation of chatbots that impersonate real people or institutions, focusing on promoting genuine interactions between users and legitimate entities.
In addition to these measures, OpenAI is actively working on enhancing transparency around the origin of information. The organization is experimenting with a provenance classifier, designed to detect images generated by Dall-e and provide users with the ability to assess images with confidence. OpenAI plans to make this tool available to a select group of testers, including journalists, platforms, and researchers, to gather feedback and improve its effectiveness.
Transparency is a key focus for OpenAI, particularly in the context of news reporting. ChatGPT, their conversational AI model, is being integrated with real-time news reporting globally. This integration will provide users with access to up-to-date news from trustworthy sources, complete with attribution and links. OpenAI believes that transparency and balanced news sources empower voters to make informed decisions and establish trust in the information they consume.
OpenAI concluded by expressing their commitment to continuing collaboration with partners and learning from their expertise to prevent any potential misuse of their tools leading up to the global elections this year. Safeguarding the democratic process and ensuring the responsible and ethical use of AI remains OpenAI’s priority.
