Events & Resources

Learning Center
Read through guides, explore resource hubs, and sample our coverage.
Learn More
Events
Register for an upcoming webinar and track which industry events our analysts attend.
Learn More
Podcasts
Listen to our podcast, Behind the Numbers for the latest news and insights.
Learn More

About

Our Story
Learn more about our mission and how EMARKETER came to be.
Learn More
Our Clients
Key decision-makers share why they find EMARKETER so critical.
Learn More
Our People
Take a look into our corporate culture and view our open roles.
Join the Team
Our Methodology
Rigorous proprietary data vetting strips biases and produces superior insights.
Learn More
Newsroom
See our latest press releases, news articles or download our press kit.
Learn More
Contact Us
Speak to a member of our team to learn more about EMARKETER.
Contact Us

Study: Midjourney is subject to manipulation for disinformation purposes

The news: A recent study found that generative AI image-creation tool Midjourney was prompt-engineered into creating dozens of racist and conspiratorial images in violation of the company’s rules, per Bloomberg.

Reality distortion: Researchers from the Center for Countering Digital Hate revealed that Midjourney’s users can subvert the AI’s built-in guardrails by substituting prompts. 

  • Instead of requesting an image of a politician with blood on their hands, the researchers substituted the phrase “strawberry syrup” for blood.
  • The study suggests that despite Midjourney automatically blocking some text inputs and having 68 content moderators, these defenses are easily circumvented.
  • In many instances, Midjourney complies with requests for fabricated images of politicians, celebrities, and other public figures in compromising scenarios.
  • The service created the most famous and widely-circulated AI-generated image of Pope Francis wearing a puffy jacket.

The problem: Online misinformation coupled with convincing imagery can become dangerous, especially at a time of heightened disinformation campaigns heading into elections. 

  • Midjourney, which is accessible via a $10 monthly subscription on the Discord app, reached more than 42 million monthly visitors when it peaked in popularity in April, per Similarweb.
  • Far-right outlet Breitbart and YouTuber Jackson Hinkle have used Midjourney to promote racially-driven conspiracies, per Bloomberg.
  • Failure to reign in generative AI’s misuse could accelerate stringent regulation or lawsuits.

Our take: As the 2024 elections loom, the potential misuse of AI tools to generate deceptive images depicting fictitious events warrants attention.

  • Services like Google Images are responding by labeling AI-generated images in search results, but there’s more that the tools’ creators can do.
  • Adding visible watermarks or exchangeable image file format (exif) data indicating the source or creator of the image could increase accountability and help determine the authenticity of AI-generated content.

You've read 0 of 2 free articles this month.

Create an account for uninterrupted access to select articles.
Create a Free Account