Feedback Is a Gift: How We Collaborate to Build Privacy-First Experiences


This week Meta hosted nearly 100 privacy academics, technologists, and thought leaders from around the world for a Global Privacy Flyout in Menlo Park to discuss how best to center privacy across product development, innovation, and tech policymaking. This is part of our ongoing work to engage outside stakeholders to embed privacy in everything we build.

I’ve been working in this space since the 90’s, when a flurry of data protection rules were enacted. Today, we are on the precipice of a new era in both policymaking and tech. Generative AI in particular has captured the world’s imagination ⏤ and regulatory attention. 

As we launch our products against this ever-evolving landscape, we rely on a global community of external stakeholders to advise us. The community’s contributions are tangible. Convenings like this week’s event are where our stakeholders advise us on design improvements to our privacy policy, user controls, and our approach to responsible AI and the metaverse. They make Meta’s products better for the billions of people who rely on them, and their work benefits the wider world.

Leadership and Accountability

To help kick off the event, the International Association of Privacy Professionals’ (IAPP) Cobun Zweifel-Keegan, Michel Protti, Meta’s Chief Officer of Privacy, Product, and I spoke about how our relationships with external stakeholders have contributed to our overhaul of privacy since 2019. We answered questions and discussed challenges regarding delivering privacy at scale, and the significant investments we’ve made to embed innovative privacy solutions across our product development lifecycle.

“Meta has come a really long way since 2019 to develop with the community, and really also  take responsibility for a lot of things, but in the end, standards and also regulation comes in when society cannot push responsibility onto one actor or several really strong actors.”

– Kirsten Rulf, Partner and Associate Director, The Boston Consulting Group 

A Global Approach

Participants traveled from nearly every continent and corner of the world: Africa, Middle East, Latin America, Australia, North America, and Asia, bringing decades of experience in data privacy. Technology may leap across borders but regulation usually does not, which makes hearing about regional variation from our partners critical to Meta’s work. 

“I do think there is enormous variation across a region and unless Meta and other organizations lean in to various national concerns, they won’t be hearing them in Menlo Park.”

– Prof. Peter Leonard, Principal, Data Synergies Pty Limited, APAC

 

“Open discussions, even when there is disagreement, it furthers knowledge and improves the technical quality of the discussion.”

– Prof. Juliano Maranhao, Director of Lawgorithm Institute and Law Professor at the University of São Paulo

 

“One thing collaboration throws up as an opportunity is to hear diverse voices. There are experiences you have that you think is your own experience, but when you share, you find out at different stages that there are people who solved that problem.”

– Gbenga Sesan, Executive Director, Paradigm Initiative, AMET

Innovative Solutions for Novel Technology

The event included several sessions on how we build protections into emerging tech. From red-team exercises to regulatory sandbox projects like Open Loop, sharing knowledge of multi-disciplinary approaches leads to better, more innovative, and more workable privacy solutions.

“At the end of the day we are think tanks, we are nonprofit organizations that are seeking to advance the use of AI and other technologies for good… let’s keep doing things that are tangible.” 

– Claudia del Pozo, Founder of EON Institute and C Minds Representative

 

“Traditional red-teaming involves you putting on the hat of a bad person trying to do a malicious thing. Now we’re like, well, why don’t you just try to be a regular person, trying to do a thing, and then getting a wrong answer? [This is] something that is very very very hard to tackle, and it’s something that needs to be tackled at scale and from a wide range of perspectives.”

– Dr. Rumman Chowdhury, Founder and CEO, Humane Intelligence

Responsible privacy design practices for emerging technologies like generative AI and AR/VR require consultation with outside experts, and we are committed to soliciting that feedback across our products.

“The type of data we’re dealing with when you deal with spatial computing devices is really novel. [Meta] was really responsive in listening to advocates and users” 

– Brittan Heller, Lecturer, Stanford University

The Work Ahead

We ended the event with the promise of many more opportunities to engage. In particular, we announced a new impact grant for our open source large language model, Llama 2. We’re working to supercharge responsible AI development by accepting applications from organizations across the globe for how Llama 2 can be applied to drive improved environmental, educational and other positive social outcomes. Submissions open October 6, and we’re looking forward to seeing what the community has to share.  

We are grateful to everyone who visited us in Menlo Park and participated in the Global Privacy Flyout. Our work to protect people’s data and build responsible products is never done, and we look forward to collaborating with this dynamic community for years to come.

You can listen to my fireside chat with my co-CPO Michel Protti and IAPP’s Cobun Zweifel-Keegan on the Privacy Conversations podcast.





Source link