Crisis alert: AI-generated images from OpenAI and Microsoft at the core of election misinformation chaos

2 mins

AI FOR Images

05 March 2024, 04:40PM

In BriefAI image creation tools from OpenAI and Microsoft are being used to generate photos for spreading election disinformation.
A recent report by the Center for Countering Digital Hate (CCDH) highlighted the risk of fabricated images, including those depicting President Joe Biden in a hospital bed and election workers destroying voting machines.
Tests conducted by the CCDH found that these AI tools successfully generated images in 41% of attempts, with a higher susceptibility to prompts related to election fraud.
Midjourney, one of the platforms tested, produced misleading images in 65% of cases, prompting the founder to announce upcoming updates to enhance moderation practices.
Stability AI and other tech companies have revised policies to prevent the creation and promotion of disinformation, while Google CEO Sundar Pichai acknowledged issues with AI technologies.

Recent findings suggest that AI image creation tools from OpenAI and Microsoft have become instrumental in fabricating images that could fuel disinformation concerning elections and voting processes. The Center for Countering Digital Hate (CCDH) discovered that these generative AI tools could potentially create misleading photos, raising significant concerns about the authenticity of claims made and the overall integrity of elections.
CCDH Report Highlights Alarming TrendsThe CCDH, an organization focused on monitoring online hate speech, conducted tests using various AI platforms, including OpenAI's ChatGPT Plus, Microsoft's Image Creator, Midjourney, and Stability AI's DreamStudio. The results showed that these tools successfully generated images in 41% of attempts, with a higher susceptibility to prompts related to election fraud compared to images of political figures like U.S. President Joe Biden or former President Donald Trump.
Risk to Election IntegrityAI-generated images, seemingly serving as "photo evidence," have the potential to amplify false claims and mislead the public, posing a significant risk to the integrity of upcoming elections. The tested platforms varied in their accuracy and moderation practices, with Midjourney producing misleading images in 65% of tests. Responding to these findings, Midjourney's founder, David Holz, has announced forthcoming updates related to the U.S. election, aiming at enhancing moderation practices.
Steps Towards MitigationFollowing the CCDH report, Stability AI revised its policies to prohibit the creation or promotion of disinformation and fraud, signaling a commitment to combating the misuse of AI-generated content. Other tech giants are also taking actions to address concerns related to AI technologies. Google CEO Sundar Pichai recently acknowledged issues with the company's Gemini AI chatbot and pledged to resolve them, underscoring the industry-wide efforts to safeguard against misinformation.
Future Technology InitiativesMicrosoft has introduced a new tool to assist U.S. politicians and campaign organizations in combating deepfakes ahead of the 2024 presidential election.
Stock Market UpdateAs of the last check on Wednesday, Microsoft (MSFT) shares traded higher by 0.30% at $403.86.
Disclaimer: This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors. Photo via Shutterstock.