The invisible biases we encode into the future of AI

Ideas & Philosophy · Published 17 October 2026 Swipe more like this →

In short: We often treat artificial intelligence as a neutral tool, but it is actually a mirror. Margaret Mitchell warns that the gaps, blind spots, and subconscious biases of its creators are being baked into the technology. To ensure AI helps rather than hurts, we must define our ethical goals today.

Margaret Mitchell, a prominent researcher in ethical AI and the current Chief Ethics Scientist at Hugging Face, argues that our current stage of artificial intelligence is merely a snapshot in a longer evolution. Because AI systems learn from human-generated data, they risk inheriting and amplifying the human biases that we often fail to recognize in ourselves. She previously served as a Staff Research Scientist at Google AI. During her time there, she co-led the company's Ethical AI group. Her academic background includes a PhD from the University of Aberdeen and a Master's from the University of Washington.

Mitchell highlights that these 'blind spots' are not just technical glitches; they are reflections of the subjective values and systemic inequalities present in our own society. To counter this, Mitchell advocates for a proactive approach to development. Rather than reacting to problems after they emerge, she emphasizes the need to define clear goals and strategies now. She has published over 100 papers on topics like sentiment classification and computer vision. Her research has been recognized by organizations like Fortune and TIME. She also holds multiple patents related to conversation generation.

Her work focuses on 'operationalizing' ethics—moving beyond abstract principles to create concrete, functional methods for responsible AI development. This includes pioneering 'Model Cards' for reporting machine learning models and developing specific methods to mitigate unwanted biases in computer vision and natural language generation. Mitchell's career reflects a commitment to using technology as a force for inclusion. Her early work at Oregon Health and Science University focused on assistive technology for non-verbal individuals and neurological disorders. By integrating diversity and inclusion into the core of computer science, she seeks to ensure that the next generation of AI serves everyone, particularly those who have been historically marginalized by technological advancement. She also developed Seeing AI to support people with low vision. Her work has earned awards from the American Foundation for the Blind. She has also received honors from Secretary of Defense Ash Carter.

General information, not professional advice. Spotted an error? Flag the card or see corrections and complaints.

Intellect Plus is a visual educational discovery platform for exploring fascinating facts and ideas. Brain Food, Not Brain Rot. What is Intellect Plus?

Related

• Are machines destined to inherit the deep-seated biases of their human creators?
• When algorithms inherit our prejudices, who is held responsible?
• When your insurance premium is decided by a biased algorithm
• Can we engineer fairness into the math of artificial intelligence?
• How do we navigate the complex ethical challenges posed by artificial intelligence bias?
• Could we survive an era of significant human unemployment?
• Are you becoming a middle manager for your own thoughts because of AI?
• AI is designing entirely new classes of antibiotics to fight superbugs
• Why increasing machine power might actually make the technology less important
• Why giving an AI a simple task could accidentally lead to catastrophe
• Is AI changing the way you should approach your next job search?
• Why teaching a robot to fetch coffee is harder than solving climate change

More in Artificial intelligence · More in Ideas & Philosophy · Latest · A–Z · All topics