Atlas of AI Risks by Nokia Bell Labs

AI is quietly shaping everyday life—deciding who gets a job, which videos we see, and even who gets flagged by facial recognition. But as AI expands, so do its risks. Could facial recognition erase privacy as we know it? Are job-matching algorithms secretly reinforcing bias? The Atlas of AI Risks, developed by Nokia Bell Labs, makes these concerns visible.

This interactive tool maps 350+ real-world AI applications, linking them to documented incidents from the news and the AI Incident Database. It classifies these applications under the EU AI Act’s risk levels and provides impact assessment cards that break down their benefits, societal concerns, and real-world incidents.

The Atlas was shaped by the public, for the public. It involved two studies with over 180 participants and consultations with legal experts. The first study explored how to effectively communicate AI risks to non-technical users, helping to refine the language, structure, and framing of the Atlas. The second study tested the Atlas with a diverse group of participants, ensuring the visualization was clear, engaging, and easy to navigate while accurately conveying risks and benefits.

Whether it’s biased job ads, AI-driven surveillance, or hidden discrimination, the Atlas of AI Risks makes the invisible, visible—giving the public the power to understand, question, and push back against AI’s growing influence.