Using GenAI to Accelerate Privacy Implementations
summary
Meta shows a concrete pipeline that uses Code Llama–based classifiers and prompt engineering to auto-annotate, review, prevent regressions in purpose‑limitation workflows. The system combines code/context signals, weighted classifier outputs and ground truth to turn high‑confidence suggestions into automated labels and diff‑time flags, integrates stack‑trace analysis to simulate flows, and catches regressions before merge. Mitigations include focused prompts, human audits, and precompute and cache strategies to control hallucination risk and API cost at Meta scale.
AI-written summary · about
chapters (6)
- 0:00Introduction
- 0:43Privacy workflows overview
- 3:06Annotating data with GenAI
- 6:26Reviewing data flows
- 8:25Preventing regressions
- 10:08Lessons learned
related talks
20mScaling Meta’s Infra with GenAI: Journey to faster and smarter Incident Response
26mAl and Software: The Good, The Bad, The Scary
28mUnlocking the power of GenAI
01h 01mAre Your Engineering Teams Prepared for the GenAI Revolution
33mOperationalizing GenAI Applications with DevOps Practices
35mOpen Source Tools To Empower Ethical and Robust AI Systems
58mIntegrating Generative AI into Modern Software Solutions
24mNatural Language Coding Revolutionising Programming with GenAI