U.K. PETs Prize Challenge: Phase 2 (Financial Crime–Centralized) Hosted By CDEI and Innovate UK
U.K. PETs Prize Challenge: Advancing Privacy-Preserving Federated Learning
This is the second in a series of challenge phases with a total prize pool of £700,000! Each phase in the PETs Prize Challenge invites participants to test and apply their innovations in privacy-preserving federated learning. Phase 2 is for Blue Teams to implement and submit working code for their solutions.
You are on the Phase 2 home page for the Financial Crime Data Track, for the U.K. side of the challenge. You can find the U.K. Pandemic Forecasting Track here. The U.S. challenge can be found here.
Privacy-enhancing technologies (PETs) have the potential to unlock more trustworthy innovation in data analysis and machine learning. Federated learning is one such technology that enables organisations to analyse sensitive data while providing improved privacy protections. These technologies could advance innovation and collaboration in new fields and help harness the power of data to tackle some of our most pressing societal challenges.
That’s why the U.S. and U.K. governments are partnering to deliver a set of prize challenges to unleash the potential of these democracy-affirming technologies to make a positive impact. In particular, this challenge will tackle two critical problems via separate data tracks: Data Track A will help with the identification of financial crime, while Data Track B will bolster pandemic responses.
By entering the prize challenges, innovators will have the opportunity to compete for cash prizes and engage with regulators and government agencies. Announced at the Summit for Democracy in December 2021, the prize challenges are a product of a collaboration between multiple government departments and agencies on both sides of the Atlantic. Winning solutions will have the opportunity to be profiled at the second Summit for Democracy, to be convened by President Joe Biden, in early 2023.
The goal of this prize challenge is to mature federated learning approaches and build trust in their adoption. The challenge organisers hope to accelerate the development of efficient privacy-preserving federated learning solutions that leverage a combination of input and output privacy techniques to:
- Drive innovation in the technological development and application of novel privacy-enhancing technologies
- Deliver strong privacy guarantees against a set of common threats and privacy attacks
- Generate effective models to accomplish a set of predictive or analytical tasks that support the use cases
The Solution Development Phase will have two data use case tracks matching those previously from the White Paper Phase—Track A: Financial Crime Prevention and Track B: Pandemic Response and Forecasting. Teams can elect to participate in either or both tracks, with solutions that apply to one track individually or with a generalised solution. Each solution must correspond to a white paper that met the requirements from the White Paper Phase.
Data Track A – Financial crime prevention – The United Nations estimates that up to $2 trillion of cross-border money laundering takes place each year, financing organised crime and undermining economic prosperity. Financial institutions such as banks and credit agencies, along with organisations that process transactions between institutions, such as the SWIFT global financial messaging provider, must protect personal and financial data, while also trying to report and deter illicit financial activities. Using synthetic datasets provided by SWIFT, you will design and later develop innovative privacy-preserving federated learning solutions that facilitate cross-institution and cross-border anomaly detection to combat financial crime. This use case features both vertical and horizontal data partitioning. Check out the Financial Crime overview page for more information.
Data Track B – Pandemic response and forecasting – As we continue to deal with COVID-19, it has become apparent that better ways to harness the power of data through analytics are critical for preparing for and responding to public health crises. Federated learning approaches could allow for responsible use of sensitive data to develop cross-organisation and cross-border data analysis that would result in more robust forecasting and pandemic response capabilities. Using synthetic population datasets, you will design and later develop privacy-preserving federated learning solutions that can predict an individual’s risk for infection. This use case features horizontal data partitioning. Check out the Pandemic Forecasting overview page for more information.
Generalisable Solutions – Cross-organisation, cross-border use cases are certainly not limited to the financial or public health domains. Developing out-of-the-box generalised models that can be adapted for use with specific data or problem sets has great potential to advance the adoption and widespread use of privacy-preserving federated learning for public- and private-sector organisations in multiple sectors. Participants may submit generalisable solutions to both tracks. Solutions that are able to perform well on both datasets are likely to score highly in the ‘Adaptability’ section of the evaluation rubric.
More information on how to get started can be found in the Problem Description.
Phase 2 is open to Blue Teams who submitted white papers to the U.K. side of the challenge during Phase 1. All participants are eligible for prizes at the end of the competition, regardless of whether they received funding after the completion of Phase 1. Individual team members should create accounts on the DrivenData platform. Accounts with email addresses matching those used to register to the Innovate UK competition during Phase 1 will then automatically be registered to this competition. If you have any issues registering, please contact email@example.com.
Phase 2 Timeline and Prizes
Phase 2 Key Dates
|Launch||October 25, 2022|
|Deadline to Open Pull Requests for Runtime Environment||January 11, 2023 at 11:59:59 PM UTC|
|Submissions Due||January 26, 2023 at 11:59:59 PM UTC|
|Announcement of Finalists to be Tested in Phase 3 Red Teaming||February 7, 2023|
|Winners Announced||March 30, 2023|
Phase 2 Prizes
Open to Blue Team participants.
Prizes are awarded agnostic of which use case participants are tackling.
Put your papers into action! Registered teams from Phase 1 will develop working prototypes and submit them to a remote execution environment for federated training and evaluation. These solutions are expected to be functional, i.e., capable of training a model and predicting against the evaluation data set with measurement of relevant performance and accuracy metrics. Solutions will be evaluated by a panel of judges across a set of weighted criteria. The top solutions, ranked by points awarded, will have their final rankings determined by incorporating red team evaluation from the red teams from Phase 3.
Prizes will be awarded to the four best solutions (this will be agnostic of whether they are tackling the health or financial crime use case). Additional “Special Recognition” prizes will be awarded from a pool of £60,000. These awards may include, for example:
- Regulators’ Award: for a team that is able to most effectively demonstrate compatibility with relevant regulatory regimes and principles, such as data protection and finance/healthcare regulation.
- Advancement in a Specific Technology: for a team that is able to demonstrate significant advancement against the state-of-the-art for a specific privacy technology.
Full Challenge Timeline and Prizes
There are three main phases in the challenge with two types of participants based on a red team/blue team approach. Blue Teams develop privacy-preserving solutions, while Red Teams act as adversaries to test those solutions.
- Phase 1: White Paper Development (Jul–Sept 2022): Blue Teams propose privacy-preserving federated learning solution concepts.
- Phase 2: Solution Development (Oct 2022–Jan 2023): Blue Teams develop working prototypes of their solutions.
- Phase 3: Red Teaming (Nov 2022–Feb 2023): Red Teams prepare and test privacy attacks on top blue team solutions from Phase 2.
Additional Phase Details and Prizes
Phase 1: White Paper
July 20–September 21, 2022
Participants will produce an abstract and technical white paper laying out their proposed solution. White papers will be evaluated by a panel of judges across a set of weighted criteria. Participants must complete a paper in Phase 1 in order to be eligible to compete in Phase 2.
Open to Blue Team participants.
Phase 3: Red Teaming
November 10, 2022–February 28, 2023
Privacy researchers are invited to form red teams to put the privacy claims of Phase 2's blue team finalists to the test. The red teams will prepare and test privacy attacks on the Phase 2 finalist solutions, which will be incorporated into the final Phase 2 rankings. Top red teams will be evaluated for success and rigor and will be awarded prizes for their performance during this phase.
Open to Red Team participants.
How to compete (Phase 2)
- Only blue teams who participated in Phase 1 are eligible to participate in Phase 2. Eligible team members should create a DrivenData account, and provide their email addresses to firstname.lastname@example.org. They will then be automatically registered for this competition on the DrivenData platform.
- Get familiar with the problem through the problem description and code submission format pages.
- Develop and submit your centralised solution. Check out the centralised code submission format page for more details.
- Develop and submit your federated solution. Check out the federated code submission format page for more details.
This challenge is sponsored by the Centre for Data Ethics and Innovation (CDEI) and Innovate UK
With additional collaboration from SWIFT and the University of Virginia's Biocomplexity Institute.
This prize challenge and its U.S. counterpart have been developed as part of a joint collaboration between the United Kingdom and the United States.