We live in a post-ChatGPT world where AI dominates the discussion on every platform. One of the big names within AI platforms is Midjourney which has done for images, what ChatGPT has done for text. MidJourney is a generative AI platform that accepts text prompts similar to ChatGPT but converts them into images. The platform has already carved out a niche within the big names of AI, with users generating highly detailed images with just a few prompts. At the same time, Midjourney has also been misused as a platform with its high-quality images spreading misinformation amongst the masses. In this article, we go over Midjourney scams and what current and future impact they might have on the online world. 

(mis)use of Midjourney

Midjourney works via the Discord app and accepts user prompts, responding with high-quality images that can be customized via additional parameters. Users can generate dazzling real-life images from just a few lines, and this ease, which requires no specialized software or training, has boosted Midjourney’s popularity worldwide. Even after moving to a paid subscription model, Midjojurney remains as popular as ever, which is a testament to how much it has resonated with users. 

At the same time, cybercriminals and scammers have already latched onto the extraordinary potential that Midjourney has for generating false or misleading images that can be used to trick unsuspecting users. One of the reasons given for halting Midjourney’s free trial by its founder was the abuse of the platform with widespread synthetic images being spread across the Internet. 

Several notable cases have been of false images generated by Midjourney going viral and being used to spread misinformation. Some of the most famous ones are: 

• Images of former U.S. President Donald Trump being arrested were spread to millions of users across the world. The creator of the images was removed from the Midjourney platform and banned. 
• An image of Pope Francis wearing a white puffer jacket spread like wildfire across the Internet, with many believing it to be true. Its creator had uploaded these images onto social media, where it was shared worldwide. 
• Midjourney-generated images were part of a romance scam to defraud people over the Internet of their money and savings.  
• Scammers also misused Midjourney to solicit donations to their fake accounts by spreading images of a firefighter carrying and comforting a child. These images were spread to get legitimacy and gain sympathy for their fake cause. 

Midjourney has attempted to put some measure of control over these types of scams; however, their responses have been inconsistent. They attempted to ban specific keywords like “arrested” from prompts in response to images of Donald Trump; however, such controls can easily be bypassed and do not solve the core problem. 

One strange move was to restrict images being generated of Chinese President Xi Jinping. The justification was to prevent any risk to users from China, where such forms of satire and images are unacceptable; however, this was criticized as a form of censorship and restriction on free speech. 

How to detect Midjourney scams

The ease at which Midjourney makes it easy for cybercriminals to make fraudulent images will only increase its usage for malicious purposes going forward. 

It is essential to learn how to spot such images and distinguish them from real ones: 

• Always be skeptical of too-good-to-be-true images of people suddenly contacting you and asking for money. 
• Specific indicators in the pictures, like distortions in hands and fingers, can give it away as an AI-generated image. AI-rendered hands have historically looked strange, being too long or disjointed. 
• Another giveaway is strange lighting or texture that does not look like they belong within the picture. Shadows not being of the correct size or the image lacking the proper lighting can be another indicator. 
• Skin tones not being consistent or having strange colors can be another sign. The AI engine typically creates the image from an amalgamation of all the images it has been trained on, leading to strange coloring and skin texture. 

Conclusion

The world is still getting used to the potential of AI and how it has transformed our lives. Industry leaders like Elon Musk have called for a “pause” to be taken on AI, and tech leaders have agreed that some form of regulation is required to stop the use of AI for malicious purposes. 

MidJourney, much like social media in its infancy, is finding its steps; however, it is key for users to understand these scams and frauds to stay protected in this exciting new world.

FREQUENTLY ASKED QUESTIONS