Centro Cultural FGV / FGV Cultural Centre | MyData track
- Ayaz Minhas, Meta Platforms, Inc.;
- Oussama Berqi, Meta Platforms, Inc.;
- Claudia Del Pozo, Eon Resilience Lab by C Minds;
- Runchao Jiang, Software Engineer, Meta;
- The use of privacy-enhancing technologies is an emerging field in the practice of privacy and is a key area to consider as companies integrate “privacy-by-design” into their products and services. The purpose of this panel is to conduct a deep dive into a number of different Privacy Preserving Machine Learning techniques (e.g. privacy adversarial attacks, differential privacy, multi-party computation) and identify which privacy “harms” they address in the Artificial Intelligence (AI) context. For example, privacy adversarial attacks demonstrate it is difficult to determine whether particular data was used to train a model, and differential privacy helps prevent reidentification of a particular data subject.
- As laws and emerging legislation around privacy and AI evolve to shift away from notice and choice and focus on technical security measures, anonymization, and risk assessments, the use of PPML and other forms of privacy enhancing technologies to address particularized privacy harms can help LATAM companies protect the data of their stakeholders and demonstrate to regulators their efforts to reduce risk. This panel will include policy and technical personnel from Meta and Eon Resilience Labs, and will also touch on Meta’s efforts to help startups and other businesses craft their own operational guidances around privacy.