Google Photos 'Gorilla' Label: AI Bias Debate Reignites

• A resurfaced 2015 incident where Google Photos labeled a Black woman as a "gorilla" has reignited debate over AI bias, especially after a commenter called it "not racist, just stupid."
• The original case involved US programmer Jacky Alcine, whose photos with a Black friend were automatically grouped into an album titled "Gorillas" by Google Photos.
• Google apologized in 2015 and "fixed" the issue by blocking the use of "gorilla" and related primate labels in Photos, a solution criticized for avoiding the root cause.
• Experts argue such misclassifications constitute "representational harm," perpetuating demeaning stereotypes regardless of intent, highlighting issues with unbalanced training data and lack of diversity in AI development.
• The incident serves as a reminder that trust in AI depends on sustained efforts to measure, disclose, and reduce inherent harms, not just advanced models.

Why It Matters: It underscores AI's racial bias, its dehumanizing impact, and the need for better solutions.