Today, Mozilla is excited to announce it has joined the Partnership on AI (PAI). PAI is a consortium of industry and civil society organizations formulating best practices for artificial intelligence — from ensuring the technology is fair, transparent, and accountable, to understanding its impact on labor, justice, and society at large.
Joining PAI was an easy decision for Mozilla. A big part of our work is fueling the movement of people building a healthier digital world: We provide fellowships to like-minded technologists and activists. We give awards to artists who are interrogating emerging technology and researchers exploring the boundaries of computing. And we freely and openly provide foundational technologies and building blocks to encourage broad innovations to improve digital life. In 2018, Mozilla took a long look at where to focus this movement-building work. We learned that as AI becomes ubiquitous, Mozilla has an opportunity to fuel the citizens, companies, technologists, governments, and organizations around the world working to make trustworthy AI a reality.
On today’s internet, users face AI that can be helpful and delightful — but also opaque and unfair. Mozilla wants to push that AI in a direction that helps, rather than harms, humanity. The stakes are high: AI is entwined with everything from privacy and fairness to safety and transparency, both online and off.
Our specific concern relates to consumer AI — how it’s built, what impact it has on us and our societies, and how innovation in AI could improve our lives. This focus includes everything from recommendation engines that influence what we see, to smart speakers that are becoming our interface to the digital world. If Mozilla and the larger movement get things right, we’ll see a world where more AI is designed with personal agency in mind, with privacy, transparency, and human wellbeing as key considerations. And, we’ll see a world where more companies are held to account when their AI systems make discriminatory decisions, abuse data, or harm people.
This isn’t new terrain for us. We’re already helping developers build more trustworthy AI, by collaborating with Pierre Omidyar and others to put $3.5 million behind professors integrating ethics into computer science curriculum. We build open-source voice technologies that power AI, including Common Voice — one of the world's largest, multi-lingual, openly-available voice data sets, built through contributions from 190,000 people in 29 languages so far. And we model what a trustworthy tech company can be, from our mission and products to how we operate, as evidenced in our commitment to lean data practices.
Now, as Mozilla deepens our investments around trustworthy AI, PAI provides us with both a place to listen and a place to speak our mind; a place to strengthen current alliances and build new ones. We’ll be working with like-minded players who are also striving to make a difference in this space, like the ACLU and Digital Asia Hub. We’ll also be able to dialogue with the people making AI technology, who we don’t always agree with: Amazon, Baidu, Facebook.
The idea isn’t to agree on everything; it’s to swap ideas, share global perspectives, and unpack what’s working and what’s not. Some of our fellow members — particularly those of the Big Nine — are creating products and practices that clash fiercely with Mozilla’s principles. By joining PAI, we’ll have another platform to reach those companies, to meet like-minded thinkers within them, and to hold them accountable.