PET Events - lfpet/pet GitHub Wiki
Linux Foundation PET Initiative BoF: Purpose and Key Themes
Purpose of the LF PET Initiative
The Linux Foundation Privacy Enhancing Technologies (PET) Initiative is a community-driven project launched to advance privacy-enhancing technologies in open source. Its core purpose is to bring together industry, academia, and open source contributors to collaborate on tools and frameworks that protect data privacy. Established under the Linux Foundation, the PET initiative is maintained by the community and governed openly – with a technical charter and contribution agreements defining its scope and governance. In essence, the project’s mission is to foster privacy-by-design in technology: enabling data to be used and shared responsibly (with minimal personal data exposure) while maintaining trust and compliance with data protection principles.
Key Themes and Focus Areas
- Open Collaboration: A key theme of the LF PET Initiative is building an open community around privacy tech. The project hosts a public mailing list and bi-weekly technical meetings to encourage broad participation and knowledge-sharing. This collaborative approach ensures that diverse stakeholders can contribute use cases, technical solutions, and best practices for privacy enhancement.
- Privacy-Preserving Data Analytics: The initiative focuses on technologies that allow valuable insights from data without compromising privacy. This includes cryptographic and architectural PETs such as secure multi-party computation, homomorphic encryption, federated learning, and confidential computing. For example, open-source frameworks like ManaTEE demonstrate how PETs and Trusted Execution Environments can enable secure analysis of sensitive data for collaborative research. By developing and integrating such tools, the PET Initiative aims to make privacy-preserving data analytics more accessible to developers and organizations.
- Security and Trust: Enhancing user trust in how data is handled is an overarching theme. PETs are designed to minimize personal data use and maximize data security, empowering individuals and organizations to control their information. The initiative supports developing standards and reference architectures that embed privacy protections at every level. This involves not only technical safeguards (encryption, anonymization, etc.) but also transparent governance, so that data use policies and privacy guarantees are verifiable and reliable. Ultimately, the LF PET community is working towards technologies that enable innovation with data while preserving confidentiality, aligning with global privacy requirements.
- Community and Education: As a Linux Foundation project, the PET Initiative also emphasizes outreach and shared learning. By maintaining open documentation and discussions, it helps educate the broader tech community about privacy-enhancing techniques. The goal is to lower the barrier to adopting PETs by providing common libraries, clear guidelines, and a forum to discuss emerging privacy challenges. This theme of community empowerment is crucial for driving adoption of PETs across different domains (from healthcare and finance to AI and cloud services).
Each of these themes – open collaboration, privacy-preserving analytics, security/trust, and community education – underpins the initiative’s efforts to make privacy-enhancing technologies accessible, effective, and widely adopted. In summary, the LF PET Initiative serves as a catalyst for developing open-source privacy tools and fostering a culture of privacy in technology.
Proposed BoF Session Title
Privacy-Enhancing Tech (PET) Initiative BoF – Community Collaboration for Data Privacy
Discussion Topics for the PET Initiative BoF Session
In this Birds of a Feather (BoF) session, participants will explore topics aligned with the Linux Foundation Privacy Enhancing Technologies (PET) Initiative’s mission and focus areas. The PET Initiative brings together industry, academia, and open-source contributors to advance privacy-by-design tools and frameworks in an open community. The discussion topics are grouped by theme to cover technical challenges, community engagement, and future directions. Each topic includes a brief description to encourage informal discussion and idea-sharing.
Technical Challenges
- Privacy-Preserving Data Analytics in Practice: How can we derive useful insights from data without compromising privacy? This topic invites discussion on implementing techniques like secure multi-party computation, homomorphic encryption, federated learning, and confidential computing in real projects. Attendees can share experiences and challenges in making privacy-preserving data analytics accessible to developers and organizations.
- Implementing Privacy-by-Design: Strategies for embedding privacy protections from the ground up in software and systems. Participants will discuss how to incorporate privacy-by-design principles – enabling data use with minimal personal exposure while maintaining trust and compliance. This is an opportunity to share best practices for building products that have privacy safeguards by default.
- Ensuring Security and Trust in Data Handling: Building user trust through strong data protection. This topic covers technical safeguards (encryption, anonymization, etc.) and how to embed privacy protections at every level of architecture. Discussion can include developing standards or reference architectures for data security and ways to verify that privacy guarantees are met, so individuals and organizations feel confident controlling their information.
Community Engagement
- Open Collaboration and Community Building: Fostering an open-source community around privacy tech. How do we encourage broad participation via public mailing lists, regular meetings, and inclusive contribution processes? Attendees can discuss ways to engage diverse stakeholders – from industry experts to academic researchers and independent developers – in contributing use cases, solutions, and best practices for privacy enhancement.
- Education and Outreach: Lowering the barrier to adopting PETs through knowledge sharing. This topic focuses on how the initiative can educate the broader tech community about privacy-enhancing techniques. Ideas might include creating accessible documentation, tutorials, or workshops, and providing common libraries or clear guidelines. The goal is to empower more developers and organizations to understand and implement PETs, driving wider adoption across various domains.
- Governance and Transparency: Ensuring the project’s community-driven nature builds trust. Discussion here revolves around the PET Initiative’s open governance model – how transparent decision-making and policies can make privacy guarantees more verifiable and reliable. Participants can share thoughts on community oversight, code of conduct, or other governance practices that help maintain an open, trustworthy collaboration environment.
Future Directions
- Emerging PET Technologies and Trends: What new privacy-enhancing technologies are on the horizon? From advances in cryptographic techniques to trusted execution environments, this topic invites brainstorming on upcoming tools or frameworks that the community should watch or contribute to. How can the PET Initiative incorporate these innovations to stay ahead of evolving privacy challenges?
- Expanding Use Cases Across Industries: Exploring opportunities to apply PET solutions in new domains. Privacy tech isn’t one-size-fits-all – how might techniques be adapted for areas like healthcare, finance, artificial intelligence, or cloud services? Participants can propose potential collaborations or pilot projects to extend privacy-preserving tech into these fields, and discuss any domain-specific requirements or challenges.
- Long-Term Vision and Roadmap: Charting the future of the PET Initiative. This topic looks at the initiative’s big-picture goals, such as developing widely adopted standards or reference architectures for privacy-by-design. What should the community prioritize moving forward to foster a culture of privacy in technology? Attendees can share ideas on setting milestones, aligning with global data protection trends, and ensuring the project’s sustainability and impact over time.
Each of these topics is intended to spark open conversation. They encourage attendees to share experiences, ask questions, and collaboratively brainstorm solutions — true to the spirit of an informal BoF session focused on privacy-enhancing technologies and open-source collaboration.
Scheduled Time and Venue
Our BoF "Privacy Enhancing Technologies" has been scheduled for Tuesday June 24 3:05pm - 3:45pm at Table 1. Unconference Sessions will take place in Bluebird Ballroom 2H. As a reminder, sessions will be set up in a round table format to encourage discussion.
Remote participation
The community can self-register as remote participants for this meeting. https://zoom-lfx.platform.linuxfoundation.org/meeting/96836698279?password=1f621762-5c9e-47fb-9f17-8af757a24a62
Join via audio
One tap mobile:
US (iOS): +12532158782,,96836698279#,,,,*489508# or +13462487799,,96836698279#,,,,*489508#
US (Android): +12532158782;96836698279#;489508# or +13462487799;96836698279#;489508#
Or dial:
US: +1 253 215 8782 or +1 346 248 7799 or +1 669 900 6833 or +1 301 715 8592
or +1 312 626 6799 or +1 646 374 8656 or 877 369 0926 (Toll Free) or 855 880 1246 (Toll Free)
Canada: +1 647 374 4685 or +1 647 558 0588 or +1 778 907 2071 or +1 204 272 7920
or +1 438 809 7799 or +1 587 328 1099 or 855 703 8985 (Toll Free)
Meeting ID: 96836698279
Meeting Passcode: 489508
International numbers: https://zoom.us/u/alwnPIaVT
Meeting materials Privacy-Enhancing Technologies (PET) Initiative (1).pdf
Participants
Anne Wu, Harvard University
Roman Zhukov, RedHat
Marcela Melara, Intel
Mike Bursell, Confidential Computing Consortium
Annie Wu, Harvard University OpenDP
Stephanie Ramirez,
Kim Berry, JP Morgan,
Kevin Zhao, Linaro
Mingshen Sun, TikTok
Ning Bao, TikTok
Tina Tsou, TikTok
Quentin Hartman, Hartkraft
And more participants we didn't catch the name.
Meeting Summary
Quick Recap The Linux Foundation's Privacy Enhancing Technology (PET) initiative held a roundtable discussion where participants were introduced to the virtual meeting environment and shared updates on their privacy initiatives, including TikTok's efforts in developing and open-sourcing privacy-preserving technologies. The meeting covered technical presentations on various privacy-focused projects, including multi-party computation and differential privacy applications, while addressing concerns about data privacy and regulatory compliance. The session concluded with discussions on GitHub repository access and collaboration processes, emphasizing the importance of community engagement and open-source contributions in advancing privacy technologies.
Next Steps Tina Tsou to update the GitHub README with links to the technical charter and other important documents. - Done Tina Tsou to fix the broken link for subscribing to the mailing list on the GitHub page. - Done Tina Tsou to add a link to the wiki from the GitHub README. - Done Tina Tsou to invite all attendees to the GitHub repository as contributors. - Done Pet Initiative team to create a Slack channel for community engagement. Pet Initiative team to prepare for an official announcement of the initiative at the EU Open Source Summit in August. Pet Initiative team to reach out to potential partners like Google, Microsoft, or Meta for collaboration on the initiative. Pet Initiative team to monitor and compile information on privacy regulations in different countries. Pet Initiative team to develop materials for educating policymakers and regulators about privacy-enhancing technologies. Summary Linux Foundation PET Initiative Roundtable:
The Linux Foundation's Privacy Enhancing Technology (PET) initiative held a roundtable discussion, with participants including Aidan from Stardos, Mike from Confidential Computing Consortium, Ning Bao from TikTok, and Tina Tsou from TikTok's influence team. The meeting focused on introductions and setting up the virtual meeting environment, with Tina Tsou providing the Zoom meeting details (ID: 963666, passcode: 489508). The session was described as being conversational in nature, with prepared slides available for reference, and was recorded for future reference.
LF PET Initiatives:
Mingshen Sun, Tina Tsou, and Ning Bao presented on LF PET initiatives, highlighting their mission to embed privacy at every layer of technology through open collaboration and the development of privacy-enhancing technologies (PETs) such as multi-party computation and homomorphic encryption. The presentation outlined three main missions: fostering open source collaboration on privacy technologies, promoting privacy-by-design principles, and improving cross-company communication about PET use cases. The initiative, which already has significant industry participation, holds bi-weekly technical meetings on Thursdays at 10 AM California time and encourages community engagement through GitHub and upcoming technical courses.
TikTok's Privacy-Preserving Technologies:
Mingshen discussed TikTok's approach to privacy-preserving technologies (PT), highlighting three key areas: enabling data-driven insights without compromising privacy through secure multi-party computation, homomorphic encryption, and federated learning; embedding strong data protection at every architecture layer using open source implementations to build trust; and community education to lower the barrier to adopting PT technologies. They detailed the Manatee project, a 2-year-old open-source initiative that provides a trusted research platform for sharing TikTok data with researchers, and mentioned that TikTok maintains both internal PT initiatives and contributes to open source projects in this space.
Privacy-Focused Ad Measurement Solution:
Tina presented a technical discussion on a privacy-focused project that combines multi-party computation (MPC) and differential privacy (DP) to address ad measurement while protecting user privacy. The approach, known as DP-PSI (Differentiation Privacy Set Intersection), allows two parties to find common items in their data without revealing individual records, using carefully calibrated noise to ensure accuracy in aggregate data without compromising individual privacy. The project, which began as a solution for TikTok's ad measurement needs, has been open-sourced to invite community collaboration and academic review, with documentation available in the TikTok Privacy Innovation repository.
Differential Privacy and Initiative Updates:
The group discussed concerns about data privacy and attacks on differential privacy techniques, with Tina explaining that while individual data is preserved in differential privacy, they approach the technology from both system design and vulnerability attack perspectives. They clarified that a privacy initiative was launched but not officially announced during the summit due to pending internal approvals, and discussed plans to announce it at the foundation open source summit in August. The team also addressed concerns about the Linux Foundation Initiative's presence on their website and GitHub, with Tina agreeing to add the Linux Foundation logo and link to the technical charter and contribution agreement in the README.
Privacy-Go: Trust Through Design:
Mingshen presented on Privacy-Go, an initiative that enables valuable insights through privacy-preserving analytics without exposing personal data. The project allows responsible data sharing and open collaboration, with the goal of protecting user data privacy while fostering innovation. Ning emphasized the importance of building trust through design and highlighted the LPA initiative as a platform for sharing knowledge and code. The presentation concluded with information on how to get involved in the PET community, including joining GitHub repositories and contributing to open-source projects.
Global Data Privacy Collaboration Initiative:
The group discussed the importance of confidentiality in computing and the need to collaborate with regulators. They agreed to monitor different regulations around the world and work together on technologies that protect privacy and data security. Tina mentioned that they would work with privacy legal colleagues to understand the regulatory landscape in various jurisdictions. The group also discussed the potential benefits of having a diverse group of stakeholders involved in these discussions with regulators.
GitHub Access Setup for Team:
The meeting focused on setting up GitHub access for team members. Tina discussed the process of inviting team members to join the GitHub repository, emphasizing the need for everyone's GitHub ID. They decided to use the traditional method of inviting members by email to ensure accuracy. The team also addressed issues with accessing the wiki and agreed to add a link to the wiki from the GitHub page.