Envisioning and Mitigating Privacy Risks for Consumer-Facing AI Product Concepts...
summary
Privy is a GenAI human-AI tool that helps non-experts draft privacy impact assessments by generating use cases, surfacing risks from an AI privacy taxonomy, and proposing concrete mitigations. A formative study with 11 practitioners shapes design and a between-subjects evaluation with 24 practitioners plus 13 independent experts rates outputs; experts find Privy LM reports higher quality than a static template, participants mark over 70% relevant and over 60% severe of identified risks, and mitigation plans score as specific and practical.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 0:53AI privacy risks
- 2:19Privid design goals
- 3:29Privid risk assessment workflow
- 6:43Privid LM vs Template study
- 7:45User study and expert findings
- 10:55Design tensions and collaboration
- 11:58Privid for product teams
- 12:40AI privacy tool goals
- 14:01Conclusion and future
related talks
11mHow Does AI Change Privacy? Towards Helping Practitioners Scope, Motivate, and Conduct
13m"We are not Future-ready": Understanding AI Privacy Risks and Existing Mitigation...
42mSecure your AI application transformation with Security
29mDeveloping Responsive Privacy Programs and Driving Impact
39mThe Search for Transparency and Accountability in the Age of...
50mPitfall or Opportunity: GenAI Legal Case Studies Revealing Practical Advice
37mFrom the Trenches Privacy Engineering Pitfalls Successes
33mPanel: How Privacy Engineers Can Shape the Coming Wave of AI Governance