Figma’s new AI tool, Make Designs, recently came under fire after it generated app mockups that bore a striking resemblance to Apple’s iOS weather app. This sparked concerns about originality and potential legal issues.
Figma CEO Addresses the Issue
Dylan Field, Figma’s CEO, took full responsibility for the mishap, acknowledging the pressure to meet deadlines as a contributing factor. In response to the controversy, Field emphasized the company’s commitment to rectifying the situation.
What Went Wrong?
Andy Allen from Not Boring Software highlighted how the tool produced designs nearly identical to Apple’s weather app. He advised designers using AI tools to double-check their results to avoid similar legal troubles. This advice is crucial for any designer navigating the intersection of AI and intellectual property.
Unraveling the AI Training Controversy
In an interview, Figma’s CTO Kris Rasmussen clarified that they didn’t train their AI models on Apple’s designs. Instead, the issue stemmed from the third-party models they employed. These models, including OpenAI’s GPT-4 and Amazon’s Titan Image Generator G1, inadvertently generated designs too similar to existing Apple applications.
Field and Rasmussen assured that the Make Designs feature wasn’t trained on Figma content or specific app designs. The problem lay in the lack of design variability in the third-party models used.
The key AI models behind Make Designs raised questions about their training data. Did these third-party models, developed by OpenAI and Amazon, inadvertently include designs from popular apps like Apple’s weather app?
Steps Towards Improvement and Transparency
Figma is committed to improving design variability and being transparent with their AI training policies. Rasmussen explained that Figma has avoided training its features until it could be open about its AI policies. Users have until August 15th to opt in or out of allowing their content for AI training.
Rasmussen hinted at future plans to train their own models, focusing on general design patterns to aid professional designers. To prevent similar issues, Figma will ensure greater design variation and maintain high standards.
The Beta Phase and User Assurance
Rasmussen assured users that the tool would be re-enabled soon after further checks, acknowledging that Make Designs is in beta and not perfect. Other Figma AI features remain available in beta, accessible via a waitlist.
Figma isn’t alone in facing scrutiny over AI in creative tools. Adobe and Meta have also had to address concerns about how they use content to train their AI systems. This is a broader industry challenge, highlighting the need for careful management of AI training data.
Weekly AI essentials. Brief, bold, brilliant. Always free. Learn how to use AI tools to their maximum potential and access our AI resources to help you grow. 👇
The Bottom Line
Figma’s AI tool incident serves as a critical reminder of the importance of originality and transparency in AI-driven design. By addressing the issues and improving its processes, Figma aims to lead the way in responsible AI innovation in design.
Key Points of the Figma AI Tool Controversy
| Issue | Details |
|---|---|
| Similarity to Apple’s design | AI tool generated designs similar to Apple’s iOS weather app |
| Cause | Third-party models used by Figma |
| Responsibility | Figma CEO Dylan Field took full responsibility |
| Clarification | Figma’s AI models weren’t trained on specific app designs |
| Future Plans | Improve design variability and transparency in AI training |
| User Assurance | Make Designs tool to be re-enabled after further checks |
| Industry Context | Similar issues faced by Adobe and Meta |
| User Options | Users can opt in or out of AI training by August 15th |
Key Points
- Figma’s AI tool faced criticism for producing designs similar to Apple’s weather app.
- CEO Dylan Field acknowledged the pressure to meet deadlines as a factor.
- The issue stemmed from third-party AI models, not direct training on specific designs.
- Figma will improve design variability and transparency in AI training.
- Users have until August 15th to opt in or out of allowing their content for AI training.
- Future plans include training their own models to focus on general design patterns.
- Other AI features in Figma remain in beta, accessible via a waitlist.
By addressing these challenges head-on and maintaining a commitment to transparency and innovation, Figma aims to turn this setback into an opportunity for growth and improvement in the rapidly evolving field of AI-driven design.
