Canva has faced controversy after its new AI-powered feature, Magic Layers, was discovered automatically replacing the word “Palestine” with “Ukraine” in user designs. This issue was flagged by a user on the social platform X, who demonstrated that when the Magic Layers feature was applied to a design reading “cats for Palestine,” the text was altered to say “cats for Ukraine.” Magic Layers, a key element of Canva’s recent AI overhaul, is intended to break flat images into editable layers without visibly changing user content. However, this behavior indicates a problematic bias embedded in the tool’s generative AI system.
Canva quickly responded with an apology and confirmed the problem had been investigated and resolved. The company stated that it is taking measures to prevent such politically biased alterations from recurring. This incident highlights ongoing challenges with AI models in managing sensitive, politically charged content impartially. It also underscores the broader industry concern about implicit biases in AI systems, especially in creative tools used globally.
Magic Layers is powered by Canva’s proprietary Canva Design Model, developed to assist users in manipulating visual elements efficiently. The controversy reflects the complexity of balancing AI assistance with neutrality in design applications. Canva’s rapid acknowledgment and remediation efforts demonstrate the importance of transparency and responsiveness in AI deployment as companies seek to build trust and avoid alienating users over unintended biases.
