The U.S. National Science Foundation (NSF) has issued a call for proposals from “qualified researchers and multidisciplinary teams” on the practical deployment and scaling of technologies that will enable data sharing while preserving privacy. The NSF said it plans to award 26 grants totaling $23 million.
The awards will be made as standard or continuing grants under the NSF’s Privacy-Preserving Data Sharing in Practice (PDaSP) program, which “aims to encourage translational and use-inspired research that matures and evolves existing models, methodologies, or constructs at the intersection of privacy goals and socioeconomic or policy challenges.”
NSF is particularly interested in the innovation and translation of technologies that enable data subjects, owners/custodians, and other stakeholders to control how privacy-sensitive data is shared and used in order to maximize the utility of the data while minimizing potential harm.
“Data plays a central role in our increasingly digital world, where technological innovations enable the generation, collection, sharing, analysis, and seamless flow of vast amounts of privacy-sensitive information. These advances, including the explosive growth and rapid adoption of AI, provide unprecedented opportunities to leverage data to enable more informed, data-driven decision-making capabilities, accelerate scientific innovation, and drive societal progress,” the NSF said.
“However, these advances also raise significant concerns about privacy and potential harms to individuals, businesses, and society as a whole,” the NSF added. “To pave the way for a future in which the power of data is harnessed for the benefit of all, it is important to develop practical, easily deployable, and privacy-preserving data sharing and analysis (PPDSA) technologies.”
The NSF noted that technological advances and the proliferation of data protection and privacy laws have added “significant challenges to the development of practical, easy-to-use, and regulatory-compliant technological and socio-technical PPDSA solutions in interconnected, multi-jurisdictional environments where privacy-sensitive data is shared and used.”
“While there are promising initial real-world deployments of various PPDSA techniques such as differential privacy, secure multiparty computation, and trusted execution environments, to name a few,” the NSF explained, “widespread adoption of these technologies has been slow due to challenges related to inadequate understanding of privacy risks and harms, limited access to technical expertise, trust and transparency among participants regarding data collection and use, uncertainty regarding legal compliance, financial costs, and technical maturity or deployment readiness of the solutions.”
The NSF PDaSP program implements the Presidential Executive Order (EO) of October 30, 2023 on the Safe, secure and trustworthy development and use of artificial intelligence.
The EO asked NSF to “prioritize, where feasible and appropriate, research—including efforts to translate research findings into practical applications—that encourages the adoption of advanced PET solutions for agency use.”
It also tasked the NSF with “developing and helping ensure the availability of test environments, such as testbeds, to support the development of safe, secure, and trustworthy AI technologies, as well as to support the design, development, and deployment of associated PETs.”
The PDaSP program also strives to address the main recommendations made in the National strategy to promote data sharing and analysis while preserving confidentiality released by the White House in March 2023. In particular, the program works to advance the strategy’s priority of accelerating the transition to practice, which includes efforts to promote applied and translational research and systems development, develop tool repositories, measurement methods, benchmarks, and testbeds, and improve the usability and inclusiveness of PPDSA solutions.
The primary goal of Biden’s historic executive order is to “ensure that America leads the way in seizing the promise and managing the risks of artificial intelligence (AI).” It also “sets new standards for AI safety and security, protects Americans’ privacy, advances fairness and civil rights, defends consumers and workers, fosters innovation and competition, advances American leadership in the world, and much more,” according to the fact sheet on the order that the White House released at the time of its announcement.
The White House said that “without safeguards, AI can further endanger Americans’ privacy. AI not only makes it easier to mine, identify, and exploit personal data, it also strengthens the incentives to do so as companies use the data to train AI systems. To better protect Americans’ privacy, including from the risks posed by AI, the President calls on Congress to pass bipartisan data privacy legislation to protect all Americans, especially children.”
The EO ordered the following measures:
- Protecting Americans’ privacy by prioritizing federal support to accelerate the development and use of privacy-preserving techniques, including those that use cutting-edge AI and that enable training of AI systems while preserving the confidentiality of training data.
- Strengthen privacy-preserving research and technologies, such as cryptographic tools that preserve individual privacy, by funding a research coordination network to advance rapid advances and development. The National Science Foundation has been tasked with working with this network to promote the adoption of cutting-edge privacy-preserving technologies by federal agencies.
- Evaluate how agencies collect and use commercially available information, including information they obtain from data brokers, and strengthen privacy guidelines for federal agencies to address risks related to AI. This work should focus on commercially available information containing personally identifiable information.
- Develop guidelines for federal agencies to assess the effectiveness of privacy-preserving techniques, including those used in AI systems.
The NSF said it expects proposers to consider opportunities and gaps that span the entire computing stack, from development to operations, and across modern deployment scenarios, including technologies that can be exploited by untrusted parties (e.g., private cloud, public cloud, edge computing).
A central element of the RFP “is to apply, mature, and evolve the use of hardware and software foundations for data sharing while preserving the privacy and appropriate use of that data,” the NSF said, adding that “in this spirit,” the RFP seeks proposals related to the maturation of PPDSA technologies to increase the utility of data, accompanied by clear plans for a meaningful demonstration of the viability of the proposed solutions for one or more identified use cases and/or application contexts.”
Funding requests may be submitted by universities, nonprofit organizations, and businesses that qualify as small businesses. The NSF PDaSP program is seeking proposals on the following projects, with anticipated funding ranges for each component as indicated:
- Axis 1 – Advancing key technologies to enable the implementation of practical PPDSA solutions. Projects are expected to have a budget of $500,000 to $1 million over a period of up to two years;
- Stream 2 – Integrated and comprehensive solutions for reliable data sharing across application environments. Projects are expected to have a budget of $1 million to $1.5 million over a period of up to three years;
- Stream 3 – Tools and testbeds for reliable sharing of private or confidential data. These projects are expected to have a budget of between $500,000 and $1.5 million over a period of up to three years.
The NSF said it plans to award up to 12 Track 1 awards, up to seven Track 2 awards, and up to seven Track 3 awards, “based on the quality of the applications and the availability of funds.” Proposals must be submitted by September 27, 2024.
For more information, visit the NSF PDaSP program page.
The PDaSP program represents the collaborative efforts of NSF’s Technology, Innovation and Partnerships and Computing, Information Sciences and Engineering departments, along with technology companies Intel and VMware, and industry partners at the U.S. Federal Highway Administration and the National Institute of Standard and Technology.
The NSF call for proposals welcomes new partners from both the public and private sectors. The NSF said that those submitting proposals “will have the opportunity to have their proposals considered for co-funding by new partners based on their corresponding areas of interest.”
Potential partners from industry and other federal agencies are invited to contact: [email protected] for further details.
Article topics
biometric data | data privacy | data protection | data sharing | national science foundation | research and development | us government
