Canva AI is biased? Company says sorry after tool replaces Palestine with Ukraine in photos
Canva has apologised after users found its AI tool replacing the word "Palestine" with "Ukraine" in designs, sparking backlash online. The company says the issue has now been fixed, and extra checks are being added to prevent similar incidents.
Canva's latest AI controversy has triggered fresh questions about how reliable generative tools really are. The popular design platform is facing criticism after users discovered that one of its new AI features was automatically replacing the word "Palestine" with "Ukraine" inside designs. What should have been a simple editing tool quickly turned into a viral debate about bias, moderation, and how AI systems make unexpected decisions.
The issue was first flagged by X user @ros_ie9, who shared screenshots showing Canva’s Magic Layers feature changing the phrase “Cats for Palestine” to “Cats for Ukraine.” The post spread quickly online, with many users questioning why the tool was altering text at all, especially without any request from the user.
Magic Layers is one of Canva's recently launched AI tools. It is designed to separate flat images into editable layers so users can modify individual parts of a design more easily. In simple terms, it should help break apart text, graphics, and elements from a finished image so they can be edited separately. It is not meant to rewrite words or replace political terms.
What made the situation stranger was that the issue appeared to be limited to the word “Palestine.” According to users who tested the feature, words such as “Gaza” were not changed. Some replies under the viral post claimed they were also able to recreate the same result before Canva rolled out a fix.
Canva apologises after what its AI tool did
As criticism grew, Canva responded with an apology and said the problem had been corrected. A company spokesperson told The Verge, "We became aware of an issue with our Magic Layers feature and moved quickly to investigate and fix it." Canva added that the issue has now been resolved and steps are being taken to stop anything similar from happening again.
The spokesperson further said, "We take reports like this very seriously, and we’re putting additional checks in place to help prevent this in future. We’re sorry for any distress this may have caused."
Canva also said the issue was isolated and did not affect designs broadly. However, the company did not fully explain how the word swap happened in the first place. It also remains unclear whether the output came from training data, internal moderation filters, or another automated system working behind the scenes.
That unanswered question is now central to the controversy. Many users are asking why a feature built to edit image layers would independently change text content. Experts often point out that AI systems can sometimes behave unpredictably because of the massive datasets and rules used to train them. Even so, when such errors touch sensitive geopolitical topics, the fallout becomes much bigger.
The timing is also awkward for Canva. The company has been aggressively expanding its AI offerings to compete more directly with Adobe and other creative software firms. Magic Layers is part of a larger AI push that Canva recently promoted as the next stage of content creation. Instead of showcasing innovation, the tool is now under scrutiny for an embarrassing mistake.
This is also not the first time an AI platform has faced allegations of bias linked to Palestine-related prompts. In past cases, other tech companies have been criticised over problematic outputs involving image generation and political language. Each new incident adds to concerns that AI systems may have hidden biases from the data they learn from or the rules they are given.
Canva's latest AI controversy has triggered fresh questions about how reliable generative tools really are. The popular design platform is facing criticism after users discovered that one of its new AI features was automatically replacing the word "Palestine" with "Ukraine" inside designs. What should have been a simple editing tool quickly turned into a viral debate about bias, moderation, and how AI systems make unexpected decisions.
The issue was first flagged by X user @ros_ie9, who shared screenshots showing Canva’s Magic Layers feature changing the phrase “Cats for Palestine” to “Cats for Ukraine.” The post spread quickly online, with many users questioning why the tool was altering text at all, especially without any request from the user.
Magic Layers is one of Canva's recently launched AI tools. It is designed to separate flat images into editable layers so users can modify individual parts of a design more easily. In simple terms, it should help break apart text, graphics, and elements from a finished image so they can be edited separately. It is not meant to rewrite words or replace political terms.
What made the situation stranger was that the issue appeared to be limited to the word “Palestine.” According to users who tested the feature, words such as “Gaza” were not changed. Some replies under the viral post claimed they were also able to recreate the same result before Canva rolled out a fix.
Canva apologises after what its AI tool did
As criticism grew, Canva responded with an apology and said the problem had been corrected. A company spokesperson told The Verge, "We became aware of an issue with our Magic Layers feature and moved quickly to investigate and fix it." Canva added that the issue has now been resolved and steps are being taken to stop anything similar from happening again.
The spokesperson further said, "We take reports like this very seriously, and we’re putting additional checks in place to help prevent this in future. We’re sorry for any distress this may have caused."
Canva also said the issue was isolated and did not affect designs broadly. However, the company did not fully explain how the word swap happened in the first place. It also remains unclear whether the output came from training data, internal moderation filters, or another automated system working behind the scenes.
That unanswered question is now central to the controversy. Many users are asking why a feature built to edit image layers would independently change text content. Experts often point out that AI systems can sometimes behave unpredictably because of the massive datasets and rules used to train them. Even so, when such errors touch sensitive geopolitical topics, the fallout becomes much bigger.
The timing is also awkward for Canva. The company has been aggressively expanding its AI offerings to compete more directly with Adobe and other creative software firms. Magic Layers is part of a larger AI push that Canva recently promoted as the next stage of content creation. Instead of showcasing innovation, the tool is now under scrutiny for an embarrassing mistake.
This is also not the first time an AI platform has faced allegations of bias linked to Palestine-related prompts. In past cases, other tech companies have been criticised over problematic outputs involving image generation and political language. Each new incident adds to concerns that AI systems may have hidden biases from the data they learn from or the rules they are given.
