Privacy For All: Mozilla’s Campaign to Think Globally, Act Locally on Privacy and AI

Today, Mozilla is launching Privacy for All — a campaign to raise data privacy standards worldwide and lay the groundwork for successful AI regulation.

Why? Privacy and AI are inextricably linked. The AI arms race has incentivized companies to scrape and amass as much data as possible for competitive advantage, with little to no regard for peoples’ privacy. For us to have trustworthy AI, we must think about privacy from a global perspective. Data knows no borders. Companies often rush to countries with lower data protection standards — especially in the Global Majority — to access the data and labor needed to train, build, and deploy systems elsewhere.

Throughout 2024, Privacy For All will work with civil society groups around the world that have spent years championing privacy protections.

The campaign follows the most recent edition of Mozilla’s *Privacy Not Included guide, which examines AI chatbots and reveals how little information is available about how these models work, including what personal data is collected and used in these systems. *PNI has investigated everything from connected cars to reproductive health apps — and gaining transparency into AI products has proven to be a particularly difficult task, as both the privacy policies and the technologies are complex and opaque. For example, in its latest guide *PNI struggled to find enough information about how the LLMs work (what data they're trained on, how they'll behave), as well as information on how personal data and users' conversations are treated (is that data part of its training by default? is opting out even possible?).

The campaign will also feature content from Mozilla campaigners, fellows, and researchers about data privacy developments across the world — from product to policy and beyond. This initiative builds on Mozilla’s decades of work standing up for online privacy, like our *PNI guide; our public policy efforts; our privacy tools like Firefox; our investments; and more.

While the campaign will engage with data protection issues around the world, it will focus strategically on passing a federal privacy law in the U.S. The U.S.’s outsized market and current legal void put millions at risk and lower how privacy is felt and experienced around the world. Many industry actors claim to support a federal privacy law, but companies’ actions reveal their priorities are to pass a weak federal standard that preempts or bypasses stronger state privacy laws. To address this, Mozilla will work with civil society organizations, researchers, technologists, and consumers to build broad support for a strong, federal privacy standard.

What does a strong privacy standard mean in practice?

Underpinning Privacy For All are 10 core privacy principles:

• Basic transparency and notice. You should know how and why companies are using your data.
• Transparency over AI. You should also know if your data is being used to train or power AI systems that make decisions about you.
• Companies should collect only the data they need to deliver the product you’ve asked for, and no more.
• Limits on third-party sharing. Companies should make it clear who they intend to share your data with, and under what circumstances — and only with your affirmative consent.
• Sensitive data like biometric or health data should receive significant protection.
• Basic data rights for consumers. People should have the right to access, delete, or correct their data.
• Reform the data broker industry. You (or someone you trust) should be able to delete your data from all data brokers at the press of a button.
• Civil rights protections. If companies can’t discriminate at a lunch counter, they shouldn’t be able to discriminate online.
• No deceptive designs, like pre-ticked boxes or tricky color schemes that make it hard to exercise your rights.
• Easy ways to express your privacy preferences once and for all, through browser opt-out signals

Privacy for All demands that everyone have key privacy rights, and that AI systems — including the datasets that fuel them — be transparent. This means that companies that use or build AI, like OpenAI or Microsoft, must be up-front about how these generative systems use individuals’ data.

Privacy For All will raise awareness about this issue — and help remedy it.

In its essence, data privacy is AI policy. The work Mozilla has invested in on Trustworthy AI over the years has taught us as much. Transparency – a core principle of data privacy – is a perhaps unglamorous but absolutely necessary step that allows regulators to investigate systems for bias and discrimination, gives researchers access to study societal harms, and empowers consumers to make informed choices. Transparency will also be key to ensuring AI is used to benefit, not harm, our communities. For 25 years, Mozilla has been on the frontlines defending your privacy. With the stakes higher than they’ve ever been, we’re doubling down until we truly have privacy for all.