Practical Ethics
summary
Mozilla presents a practical policy: lean data practices — collect only what you need, keep it for the minimum time, and avoid violating user expectations. Data is split into four classes of data: technical (opt-out, low risk), interaction (opt-out, distributional metrics), web activity (URLs — opt-out, rare, with mitigation because reidentification and side effects are common) and highly sensitive (generally not collected; opt‑in with explicit consent). Documentation and guidelines live on GitHub.
AI-written summary · about
chapters (8)
- 0:00How To Be Perfect
- 0:32Practical Ethics
- 1:21Collect only what you need Keep it for the minimum amount of time
- 1:49Classes of Data
- 2:05Technical Data Examples: Os, available memory, version number Generally okay to colect opt-out
- 2:29Interaction Data Examples of tabe, session length, config settings, feature se Generally okay to colect, opt-out.
- 3:04Web Activity Data Example: browsing history Slickier. Usually no, but may be possible with migation
- 4:19Highly Sensitive Data Examples: email, username, identifiers Assume no Maybe opt-in with advance notice, user consent, and secondary opt-out
related talks
28mWhy Mozilla is focussing on better data
37mThe balance between data privacy and user experience
47mDesigning for User-Centered Privacy
39mQ&A session -- The Big Data Paradox public lecture
20mThe Untold Benefits of Ethical Design
32mThe Ethical Choice
01h 14mEthical data handling
11mThe Future of Privacy