OpenAI’s strategy to prevent election misinformation in 2024: Restraining AI’s influence
OpenAI, the creator of ChatGPT, has announced measures to curb the dissemination of election misinformation through its AI tools, as numerous countries gear up for national elections. In a recent blog post, the San Francisco-based AI startup outlined a range of safeguards, combining existing policies with new initiatives, to counter the potential misuse of its highly popular generative AI tools. While these tools can generate text and images rapidly, they also have the potential to be exploited for the creation of deceptive content, including misleading messages and realistic fake photographs.
The steps apply specifically to OpenAI, just one player in the expanding field of companies developing advanced generative AI tools. The company, which announced the move on Monday, said it plans to “continue our platform security efforts by adding accurate voting data, monitoring measured practices and improving transparency.”
It said it prohibits people from using its technology to create chatbots that impersonate real candidates or governments, distort how voting works or prevent people from voting. It said that until the persuasive power of its technology can be studied more, it will not allow users to build apps for political campaigning or lobbying.
OpenAI said that “starting early this year” it will digitally watermark AI images created using the DALL-E image generator. This permanently marks the content with its origin information, making it easier to identify whether an image seen elsewhere on the web was created by an AI tool.
The company also said it is working with the National Association of State Secretaries to direct ChatGPT users with logistical questions about voting to get accurate information on that group’s nonpartisan website, CanIVote.org.
Mekela Panditharatne, an adviser to the democracy program at the Brennan Center for Justice, said OpenAI’s plans are a positive step toward combating election fraud, but it depends on how they are implemented.
“For example, how exhaustive and comprehensive are the filters when flagging questions about the election process?” he said. “Do objects slip through the cracks?”
OpenAI’s ChatGPT and DALL-E are the most powerful creative AI tools to date. But there are many companies with equally sophisticated technology that don’t have as many safeguards against election disinformation.
While some social media companies, such as YouTube and Meta, have adopted AI flagging policies, it remains to be seen whether they can consistently catch infringers.
“It would be helpful if other generative AI companies adopted similar guidelines to enforce rules of practice across the industry,” said Darrell West, a senior fellow at the Brooking Institution’s Center for Technology Innovation.
Without voluntary industry-wide adoption of such policies, regulating AI-generated disinformation in politics would require legislation. In the United States, Congress has yet to pass legislation seeking to regulate the industry’s role in politics, despite bipartisan support. Meanwhile, more than a third of US states have passed or introduced bills to crack down on fraud in political campaigns as federal legislation stalled.
OpenAI CEO Sam Altman said that even with all of his company’s safeguards in place, his mind is not at ease.
“I think it’s good that we have a lot of anxiety and we’re doing everything we can to get it as good as possible,” he said in an interview Tuesday at a Bloomberg event during the World Economic Forum in Davos, Switzerland. “We have to monitor this incredibly closely this year. Super tight monitoring. Super tight feedback loop.”
Also read these top stories today:
Apple vs Epic Games! The CEO of Fortnite maker Epic Games said Tuesday that the company’s legal battle to open up Apple’s iPhone to alternative app stores has been lost after the US Supreme Court declined to hear the case Know What Happened Here. If you enjoyed reading this article, please share it with your friends and family.
AI Fallout in the financial world! Major language models like OpenAI’s ChatGPT threaten to disrupt most areas of life and work. Financial trading is no exception. Dive here. If you enjoyed reading this article, please share it with your friends and family.
An opportunity for the founder of Facebook? The new $3,500 Apple Vision Pro headphones are fueling growing interest in the metaverse. How can Mark Zuckerberg capitalize? Fitness. Jump in here.