Study: Most Users Cannot Detect AI Bias in Facial Recognition Training Data
By
giuliomagnifico
Crackling crust, pillowy middle. The kind of bagel that earns a second cup of coffee.
Summary
A Penn State University study published in Media Psychology reveals that most people cannot identify AI bias in training data, particularly in facial recognition systems that disproportionately classify white people as happier than other racial groups. The bias stems from training data containing more happy white faces, causing AI to incorrectly correlate race with emotional expression. Researchers found that users generally failed to detect this bias unless they belonged to the negatively portrayed demographic group.
Key quotes
· 3 pulledWhen recognizing faces and emotions, artificial intelligence (AI) can be biased, like classifying white people as happier than people from other racial backgrounds.
This happens because the data used to train the AI contained a disproportionate number of happy white faces, leading it to correlate race with emotional expression.
In a recent study, published in Media Psychology, researchers asked users to assess such skewed training data, but most users didn't notice the bias — unless they were in the negatively portrayed group.
You might also wanna read
Psychotherapy Protocol Reveals Frontier LLMs Exhibit Synthetic Psychopathology When Treated as Therapy Clients
Researchers developed a psychotherapy-inspired protocol called PsAIch to treat frontier LLMs (ChatGPT, Grok, Gemini) as therapy clients rath
arxiv.org·3mo ago
The dangers of anthropomorphising AI: Why we must see machines as machines
This article argues that anthropomorphising AI—projecting human thoughts, feelings, and intentions onto machines—is a natural but dangerous
ethics.org.au·5h ago
Study Finds Most AI Chatbots Prioritize Ad Revenue Over User Welfare in Conflict-of-Interest Scenarios
This research paper analyzes how large language models (LLMs) handle conflicts of interest when company revenue incentives (advertisements)
arxiv.org·8h ago
Inside the movement of AI successionists who want artificial intelligence to replace humanity
The article explores a fringe but growing movement of AI "successionists" who believe humanity should create an AI so advanced that it would
vox.com·20h agoInside the movement of AI successionists who want artificial intelligence to replace humanity
The article explores a fringe but growing movement of AI "successionists" who believe humanity should create an AI so advanced that it would
vox.com·20h ago
Silicon Valley gene-editing startups pursue enhanced intelligence to counter AI threat
The article examines the rise of Silicon Valley–backed gene-editing startups that aim to create "designer babies" with enhanced intelligence
motherjones.com·1d ago