OpenAI says working to prevent abuse, provide transparency on AI-generated content, and improve access to accurate voting information
In a major development to ensure election integrity across the globe, the world’s leading artificial intelligence (AI) firm OpenAI has announced to launch tools to counter disinformation.
The ChatGPT maker, in a blog post, stated that the company is “working to prevent abuse, provide transparency on AI-generated content, and improve access to accurate voting information”.
The move is considered significant amid the dangers of fake news and misinformation affecting the electoral processes around the world with a slew of countries going to polls including Pakistan, India, US and the European Union.
The World Economic Forum (WEF) in its Global Risk Report has also declared AI-driven misinformation as the “biggest short-term threat” to the global economy.
Vowing to stop the harmful use of its technology — ChatGPT and DALL·E, OpenAI sought to invite all stakeholders to protect the integrity of elections.
“We want to make sure our technology is not used in a way that could undermine this process. We want to make sure that our AI systems are built, deployed, and used safely. Like any new technology, these tools come with benefits and challenges. They are also unprecedented, and we will keep evolving our approach as we learn more about how our tools are used,” the company said.
Preventing abuse
The company added that before releasing new systems, they red team them, engage users and external partners for feedback, and build safety mitigations to reduce the potential for harm.
“DALL-E has guardrails to decline requests that ask for image generation of real people, including candidates,” it said.
OpenAI said they are still working to understand how effective their tools might be for personalised persuasion and until it is clear they won’t allow people to build applications for political campaigning and lobbying.
The company has also built new GPTs through which users can report potential violations.
Additionally, a mechanism has been placed that wouldn’t let builders create chatbots that pretend to be real people or institutions.
Likewise, applications that can deter people from participating in democratic processes won’t be allowed.
Transparency around content
Noting the importance of labelling AI-crafted content, OpenAI said the company is working on several provenance efforts that would attach reliable attribution to the text generated by ChatGPT, and also give users the ability to detect if an image was created using DALL-E 3.
“Early this year, we will implement the Coalition for Content Provenance and Authenticity’s digital credentials — an approach that encodes details about the content’s provenance using cryptography — for images generated by DALL-E 3.”
The coalition, according to AFP, also known as C2PA, aims to improve methods for identifying and tracing digital content. Its members include Microsoft, Sony, Adobe and Japanese imaging firms Nikon and Canon.
Collaboration
ChatGPT, the statement added, is increasingly integrating with existing sources of information after which users will start to get access to real-time news reporting globally, including attribution and links.
Similarly, to improve access to authoritative voting information, OpenAI has joined hands with the National Association of Secretaries of State (NASS) where ChatGPT will direct users to CanIVote.org — an authoritative website on US voting information.
