In recent weeks, users of the Leonardo AI platform have raised concerns about unexpected watermarking on their generated assets. This issue has stirred discussions among digital artists, content creators, and developers who rely on Leonardo AI for high-quality, unbranded visuals. These concerns were further amplified by a sudden change in asset attribution settings, which reset user preferences and quietly removed Leonardo’s auto-branding feature.
TL;DR
Leonardo AI users started noticing unexpected watermarks on generated images due to a glitch or undisclosed update. Around the same time, the company pushed a reset on asset attribution settings, removing an auto-branding feature that formerly appended credits. While the platform seems to be stabilizing the issue, many users remain confused and seek transparency from the Leonardo team. A fix is underway, with some clarity provided through official channels.
The Emergence of Unexpected Watermarks
Leonardo AI, a popular generative platform used for creating art, character designs, and marketing visuals, has always allowed users to generate clean imagery based on specific style models. However, in early 2024, users began to report that images suddenly included visible watermarks, even when such marks had not been manually activated.
The watermarks—often subtle but noticeable patterns or branding emblems—sparked confusion. Was Leonardo reintroducing a branding requirement? Or was this simply a bug introduced in the latest update?
Many users turned to online communities like Reddit, Discord, and X (formerly Twitter) to share impacted files, speculate on causes, and demand official clarification. Some believed it to be a new policy aimed at increasing platform attribution, possibly driven by licensing reasons. Others suspected coding errors or server misconfigurations that affected asset rendering logic.
[p ai-img]watermark image, ai art example, digital illustration affected[/ai-img]
The Role of Asset Attribution Settings
Concurrent with the watermark confusion was another disruptive change: asset attribution preferences were suddenly reset. Users who had long disabled auto-branding—an optional feature that appends “Generated by Leonardo AI” to images and metadata—found the setting returned to default.
This reset was not immediately announced to the community, leading to suspicion and frustration. Key impacts included:
- Unwanted credit tags appearing in exported files
- Loss of user customization for attribution
- Concerns around ownership and distribution rights with branded content
Creators who distribute through stock marketplaces or client workspaces found themselves forced to reprocess imagery to remove undesired branding. For professionals, even a minor attribution can be perceived as undesirable or unprofessional—especially when clean output had been the norm before.
Leonardo AI’s Response to the Outcry
Amid the growing criticism, Leonardo AI acknowledged the concerns via a post on their official help desk and Discord server channels. The team explained that the sudden watermarks were not intentional and were due to a rendering bug linked to a backend server update. The bug inadvertently toggled watermarking on some image generations even when disabled in project settings.
The attribution reset, on the other hand, was the result of a policy-wide system upgrade meant to synchronize terms of service with new licensing frameworks. In their statement, Leonardo clarified:
“We apologize for the confusion caused by the attribution resets. These were not meant to override user trust, and we are working to restore prior settings. In the future, we aim to roll out these changes with in-app announcements for better clarity.”
To address immediate user concerns, Leonardo has initiated the following actions:
- Hotfixes to rendering API endpoints to prevent unintentional watermark additions
- User alerts within the app outlining attribution policy changes
- Option to toggle auto-branding OFF/ON on a per-image basis
- A support form allowing users to request re-exports of prior assets without watermarks
While these actions have helped stabilize some of the workflows, many frequent users argue that trust has been eroded. One major demand is for greater changelog transparency and a roadmap to accompany feature rollouts.
[h3]Understanding Why Attribution and Branding Matter to Creators[/h3]
In the digital content economy, control over branding and attribution is paramount. When users commission AI-generated works or use such tools for business, unsolicited signatures or watermarks can:
- Compromise exclusivity in projects meant for clients or corporate branding
- Raise licensing ambiguities when submitted to publishing, advertising, or commercial sales platforms
- Undermine aesthetic choices where branding elements do not fit the visual tone
Leonardo AI had earned praise initially for its flexible export options, letting users manage attribution based on their use-case or licensing tier. Therefore, the sudden override felt regressive to many professionals.
[p ai-img]creative workspace, graphic designer, ai art creation tools[/ai-img]
Steps Moving Forward
To prevent future disruptions, developers at Leonardo are reportedly working on a more robust settings management system. This will include:
- Persistent user preferences stored securely across updates
- A versioned changelog visible from the interface whenever policy or setting changes occur
- Clear user prompts when any AI-generated output includes branding, attribution metadata, or embedded watermarks
While no ETA has been provided yet, Leonardo’s spokesperson emphasized that these changes are being prioritized based on user feedback and retention concerns. In a competitive landscape populated by rivals like Midjourney, DALL·E, and Stable Diffusion, maintaining trust is essential for longevity.
In the meantime, power users are sharing community workarounds. For example, some have designed drag-and-drop watermarked image cleaners, while others revert to older API endpoints known to bypass watermark flags entirely (where permitted by their TOS agreement).
What This Means for Leonardo AI’s Future
This particular incident highlights the thin line generative platforms must walk: balancing attribution and identity without alienating committed professional users. Leonardo AI’s usefulness hinges not just on technical capability but on how intuitive and transparent it feels to creators.
As AI continues to shape media, art, and commerce, expect attribution policies to remain a hot topic. User empowerment, licensing control, and watermark transparency will all be factors that define the leaders in this emerging space.
Frequently Asked Questions (FAQ)
- What caused the unexpected watermarks?
- This was due to a rendering glitch introduced during a backend upgrade. The watermarking was unintentionally activated on some outputs even when toggled off by users.
- Why were attribution settings reset without notice?
- Leonardo AI initiated a system-wide update that synced attribution policies with new licensing standards, but failed to alert users in-app. This caused all auto-branding preferences to revert to default.
- Can I remove watermarks from past images?
- Yes. Users can submit a request via Leonardo’s support portal to have watermarked images re-exported without branding, depending on license terms and generation history.
- Will settings remain persistent in the future?
- Leonardo AI has stated that they are enhancing their settings infrastructure to ensure user preferences persist through updates and system patches.
- Is watermarking permanent for all users?
- No. Watermarking remains optional unless required by specific subscription plans or licensing terms. Users should double-check their current project settings for confirmation.
