25 Sep 2025 1 min read

Developers should not have access to eye tracking information

...and there are strong reasons against it. The conversation could be philosophical, but it highlights the conflicting opinions on whether commercial interests should take precedence over human well-being. Also, the few potential exceptions could be handled in controlled environments using entitlements or research.

“This figure illustrates the types of data commonly captured by eye trackers and the potential personal information that can be inferred. On the left, several eye-related biometric features are listed, including eye opening and closure (e.g., blink duration), eye movements (e.g., saccades, fixations), eye status (e.g., redness), pupil properties (e.g., size, reactivity), iris characteristics (e.g., color, texture), and facial attributes (e.g., wrinkles, expressions). These features are visually linked to a central box labeled ‘possible inference of personal information,’ which in turn connects to a range of possible inferences shown on the right. These inferences include gender, age, geographical origin, biometric identity, physical health (such as vision disorders, pain, or Parkinson’s disease), cultural background, mental health (such as depression, PTSD, autism, eating disorders), personality traits (e.g., extroversion, neuroticism), skills and abilities (such as math, sports, chess), level of sleepiness, drug consumption (alcohol, tobacco, cocaine, etc.), mental workload, and cognitive processes (expectations, reasoning, memory). The diagram summarizes how eye tracking data can reveal diverse sensitive and personal information.” — Overview of sensitive inferences that can be drawn from eye tracking data.