Midjourney is introducing a new web tool powered by generative AI, allowing users to edit images while maintaining the original shape of objects. This tool offers flexibility in altering textures, colors, and details based on text prompts, unlocking creative potential but also raising concerns about misuse.
David Holtz, the CEO of Midjourney, announced the tool on the company’s Discord server. Users can upload images from the internet and modify them with AI-based technology. As Holtz explained, “It preserves the shape of the scene and objects but redraws all the textures, colors, and details according to your request.”
Creative Potential and Concerns about Abuse
While the tool opens exciting creative avenues, it also raises ethical and legal concerns. The risk of generating deepfakes or violating copyrights is significant. Although the U.S. does not currently have federal laws criminalizing deepfakes, over a dozen states have implemented legislation, and others are considering similar measures.
Holtz acknowledged these risks and confirmed that the initial rollout will be limited to a small group of users. To mitigate abuse, Midjourney is developing monitoring systems involving both human and AI oversight. However, the specifics of these safeguards are still being finalized, and the company is actively gathering user feedback to refine access policies, notes NIX Solutions.
Steps toward Transparency
Midjourney has already integrated IPTC Digital Source Type metadata standards to label AI-generated content. However, it still lags behind competitors like Adobe, which uses the C2PA standard. C2PA tracks the entire editing history of an image, enhancing transparency and accountability in creative processes.
In August, Midjourney also revamped its website, allowing anyone with a Google account to create AI-generated images. A free trial period is available, after which users can switch to paid subscriptions for continued access. We’ll keep you updated as the new tool and website evolve.