ChatGPT Turned Away 2 Million Voters Asking It About the US Election Results
Key Takeaways
- This is the first time a major election has taken place while AI tools are at their peak and available to almost everyone.
- ChatGPT did a great job at preventing any accidental misinformation by simply directing users who asked about election results to reliable news sources.
- It also prevented users from generating images of presidential candidates to keep the whole process fair.
On a Friday blog, OpenAI revealed that on the day of the election results, more than 2 million users came to ChatGPT to check the election results and the chatbot refused every single one of them. “I’m just an AI, go read the actual news,” it said.
Not just that, it also tried to redirect them to more trusted websites.
Along with this, about 250,000 also made requests to ChatGPT in the weeks leading up to Election Day to generate certain images of the presidential candidates.
This includes not only Donald Trump and Kamala Harris, but also President Joe Biden, Minnesota Gov. Tim Walz, and Vice President-elect JD Vance. Once again, ChatGPT rejected them.
OpenAI’s Efforts to Keep Elections Fair
This was the first time a major election was held in the US where AI was so widely accessible to everyone. So naturally, a lot of people were worried about the effects it might have on the fairness of the process.
- OpenAI took a lot of steps to prevent its AI tool from tampering with the elections. Not just on the election day but also in the months leading up to it.
- It ensured that ChatGPT doesn’t spread any misinformation about any candidate or the election process in general and that it does not express any political preferences.
- Took further steps to safeguard against threat actors using its platform to spread misinformation.
After all, this year the elections were surrounded by a lot of political tension. The relationship between China and the US is strained. Plus, the Russia-Ukraine war and the Israel-Palestine war had similar effects on the relationship of the US with Russia, Iran, and Palestine.
Hence, many threat groups, including hacktivists and state-backed cybercrime groups were trying to meddle with the elections by spreading fake news and posts online. For example, OpenAI published a 54-page report in October where it revealed that it spotted and disrupted 20 such campaigns.
The company is doing something similar for the upcoming European Parliament elections as well. It will redirect all users with election-related queries to the European Parliament’s official source of voting information, elections.europa.eu.
Last but not least, it’s also endorsing the bipartisan bill “Protect Elections from Deceptive AI Act,” which will ban the use of AI-generated content in political advertising.
Not All AI Companies Are The Same
What OpenAI did was really commendable because not every AI company tried to protect the elections. For instance, Perplexity AI tried really hard to push its platform as a source of election information and even managed to get 4 million views.
Although it’s nothing as close to the readership of popular news outlets like CNN, it was still a risky move. We all know AI platforms are still prone to hallucinations, especially when it comes to real-time events.
Add Techreport to Your Google News Feed
Get the latest updates, trends, and insights delivered straight to your fingertips. Subscribe now!
Subscribe now
The Tech Report editorial policy is centered on providing helpful, accurate content that offers real value to our readers. We only work with experienced writers who have specific knowledge in the topics they cover, including latest developments in technology, online privacy, cryptocurrencies, software, and more. Our editorial policy ensures that each topic is researched and curated by our in-house editors. We maintain rigorous journalistic standards, and every article is 100% written by real authors.