Midjourney 6 raises concerns with its hyper-realistic fake images

Realistically fake: San Francisco-based AI startup Midjourney has ignited concerns over the potential misuse of its sophisticated AI image generator, which can create hyper-realistic images of individuals, including prominent political figures. The company's latest version, Midjourney 6, boasts of capabilities that can produce lifelike images, raising questions about the proliferation of misinformation and propaganda.

Founded by David Holz, who previously co-founded Leap Motion, Midjourney has gained major traction despite operating under the radar without venture capital backing or marketing efforts. With an estimated 9 million users on its Discord server, Midjourney's AI has the capacity to create images so realistic that they blur the lines between fact and fiction.

Midjourney generated images of Narendra Modi and Donald Trump

Holz has previously expressed discomfort with the potential for generating fake images but his pursuit of improving Midjourney's capabilities day-by-day has made it capable of misuse too. 

The implications of Midjourney's technology extend beyond mere aesthetic concerns. Concerns have emerged about the use of AI-generated images in propagandistic contexts, potentially influencing public opinion through deceptive means. For example, images placing political figures like Donald Trump and Prime Minister Narendra Modi in scenarios which can harm their political image. While Midjourney has taken steps to monitor its platform, including employing approximately 70 staff members to oversee content, challenges persist.

Midjourney generated an image of Donald Trump being arrested. 

Restrictions designed to prevent the creation of harmful or misleading content can be circumvented through different images. 

Midjourney's technology underscores broader debates about the ethical responsibilities of AI developers and the need for robust safeguards against the proliferation of misinformation.