Randomness Concerns when Deploying Differential Privacy
Open accessAbstract
The U.S. Census Bureau is using differential privacy (DP) to protect confidential respondent data collected for the 2020 Decennial Census of Population & Housing. The Census Bureau's DP system is implemented in the Disclosure Avoidance System (DAS) and requires a source of random numbers. We estimate that the 2020 Census will require roughly 90TB of random bytes to protect the person and household tables. Although there are critical differences between cryptography and DP, they have similar requirements for randomness. We review the history of random number generation on deterministic computersømitt, including von Neumann's "middle-square'' method, Mersenne Twister (MT19937) (the default NumPy random number generator, which we conclude is unacceptable for use in production privacy-preserving systems), and the Linux /dev/urandom device. We also review hardware random number generator schemes, including the use of so-called "Lava Lamps'' and the Intel Secure Key RDRAND instruction. We finally present our plan for generating random bits in the Amazon Web Services (AWS) environment using AES-CTR-DRBG seeded by mixing bits from/dev/urandom and the Intel Secure Key RDSEED instruction, a compromise of our desire to rely on a trusted hardware implementation, the unease of our external reviewers in trusting a hardware-only implementation, and the need to generate so many random bits.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Cryptographically secure random noise generation is crucial to avoid vulnerabilities associated with default pseudo-random number generators.
“Moreover, our framework includes support for cryptographically secure random noise generation which is crucial to avoid vulnerabilities associated with default pseudo-random number generators33.”
From Medical imaging deep learning with differential privacy (2021) · cites this paper as [33]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
Calibrating Noise to Sensitivity in Private Data Analysis
Dwork, C., McSherry, F., Nissim, K., & Smith, A. (2006) · Lecture Notes in Computer Science · doi:10.1007/11681878_14 · Cited by 7,272 · see citing works
National Conference of State Legislatures
Clink, K. (2011) · Reference Reviews · doi:10.1108/09504121111114072 · Cited by 2,321 · see citing works
Functional Encryption: Definitions and Challenges
Boneh, D., Sahai, A., & Waters, B. R. (2011) · Lecture notes in computer science · doi:10.1007/978-3-642-19571-6_16 · Cited by 1,015 · see citing works
TestU01: A C library for empirical testing of random number generators
L'Ecuyer, P., & Simard, R. (2007) · ACM Transactions on Mathematical Software · doi:10.1145/1268776.1268777 · Cited by 1,014 · see citing works
Inference Attacks on Property-Preserving Encrypted Databases
Naveed, M., Kamara, S., & Wright, C. V. (2015) · ACM SIGSAC Conference on Computer and Communications Security (CCS) · doi:10.1145/2810103.2813651 · Cited by 511 · see citing works
The U.S. Census Bureau Adopts Differential Privacy
Abowd, J. M. (2018) · ACM SIGKDD International Conference on Knowledge Discovery & Data Mining (KDD) · doi:10.1145/3219819.3226070 · Cited by 294 · see citing works
Universally utility-maximizing privacy mechanisms
Ghosh, A., Roughgarden, T., & Sundararajan, M. (2009) · ACM Symposium on Theory of Computing (STOC) · doi:10.1145/1536414.1536464 · Cited by 247 · see citing works
Randomness Requirements for Security
Eastlake, D., Schiller, J. I., & Crocker, S. D. (2005) · doi:10.17487/rfc4086 · Cited by 239 · see citing works
Computational Differential Privacy
Mironov, I., Pandey, O., Reingold, O., & Vadhan, S. (2009) · Lecture notes in computer science · doi:10.1007/978-3-642-03356-8_8 · Cited by 234 · see citing works
On significance of the least significant bits for differential privacy
Mironov, I. (2012) · ACM SIGSAC Conference on Computer and Communications Security (CCS) · doi:10.1145/2382196.2382264 · Cited by 184 · see citing works
ChaCha20 and Poly1305 for IETF Protocols
Nir, Y., & Langley, A. (2018) · doi:10.17487/rfc8439 · Cited by 178 · see citing works
Issues Encountered Deploying Differential Privacy
Garfinkel, S. L., Abowd, J. M., & Powazek, S. (2018) · Proceedings of the 2018 Workshop on Privacy in the Electronic Society · doi:10.1145/3267323.3268949 · Cited by 112 · see citing works
Self-citationWhat Else is Revealed by Order-Revealing Encryption?
Durak, F. B., DuBuisson, T. M., & Cash, D. (2016) · ACM SIGSAC Conference on Computer and Communications Security (CCS) · doi:10.1145/2976749.2978379 · Cited by 108 · see citing works
Differential Privacy and Census Data: Implications for Social and Economic Research
Ruggles, S., Fitch, C. A., Magnuson, D. L., & Schroeder, J. (2019) · AEA Papers and Proceedings · doi:10.1257/pandp.20191107 · Cited by 91 · see citing works
Optimizing error of high-dimensional statistical queries under differential privacy
McKenna, R., Miklau, G., Hay, M., & Machanavajjhala, A. (2018) · Proceedings of the VLDB Endowment · doi:10.14778/3231751.3231769 · Cited by 84 · see citing works
Gennaro, R. (2006) · IEEE Security & Privacy · doi:10.1109/msp.2006.49 · Cited by 60 · see citing works
Information technology in the laboratory
Mahaffey, R. R. (1991) · Chemometrics and Intelligent Laboratory Systems · doi:10.1016/0169-7439(91)90022-t · Cited by 38 · see citing works
Implementing Differential Privacy: Seven Lessons From the 2020 United States Census
Hawes, M. B. (2020) · Harvard Data Science Review · doi:10.1162/99608f92.353c6f99 · Cited by 32 · see citing works
Differential Privacy and the US Census
Dwork, C. (2019) · ACM SIGMOD-SIGACT-SIGAI Symposium on Principles of Database Systems (PODS) · doi:10.1145/3294052.3322188 · Cited by 29 · see citing works
A Provable-Security Analysis of Intel’s Secure Key RNG
Shrimpton, T., & Terashima, R. S. (2015) · Lecture notes in computer science · doi:10.1007/978-3-662-46800-5_4 · Cited by 29 · see citing works
Preserving differential privacy under finite-precision semantics
Gazeau, I., Miller, D., & Palamidessi, C. (2016) · Theoretical Computer Science · doi:10.1016/j.tcs.2016.01.015 · Cited by 26 · see citing works
Differential Privacy with Imperfect Randomness
Dodis, Y., López-Alt, A., Mironov, I., & Vadhan, S. (2012) · Lecture notes in computer science · doi:10.1007/978-3-642-32009-5_29 · Cited by 26 · see citing works
SoK: Security Models for Pseudo-Random Number Generators
Ruhault, S. (2017) · IACR Transactions on Symmetric Cryptology · doi:10.46586/tosc.v2017.i1.506-544 · Cited by 25 · see citing works
Practical State Recovery Attacks against Legacy RNG Implementations
Cohney, S., Green, M., & Heninger, N. (2018) · ACM SIGSAC Conference on Computer and Communications Security (CCS) · doi:10.1145/3243734.3243756 · Cited by 19 · see citing works
SoK: Security Models for Pseudo-Random Number Generators
Ruhault, S. (2017) · DOAJ (DOAJ: Directory of Open Access Journals) · doi:10.13154/tosc.v2017.i1.506-544 · Cited by 14 · see citing works
Leveraging Platform Weaknesses
Shema, M. (2012) · Elsevier eBooks · doi:10.1016/b978-1-59-749951-4.00007-2 · Cited by 4 · see citing works
Flagged first, then by citations
Research works printed but not linked: 13
- S. Das, J. Werner, M. Antonakakis, M. Polychronakis, and F. Monrose. 2019. SoK: The Challenges, Pitfalls, and Perils of Using Hardware Performance Counters for Security. In 2019 IEEE Symposium on Security and Privacy (SP). 20-38.
- Cynthia Dwork, Frank McSherry, Kobbi Nissim, and Adam Smith. 2006. Cali- brating Noise to Sensitivity in Private Data Analysis. In Proceedings of the Third Conference on Theory of Cryptography (TCCâĂŹ06). Springer-Verlag, Berlin, Hei- delberg, 265âĂŞ284. https:// doi.org/ 10.1007/ 11681878_14 · DOI · Paper page
- David Alan Grier. 2007. When Computers Were Human. Princeton University Press, USA.
- Z. Gutterman, B. Pinkas, and T. Reinman. 2006. Analysis of the Linux random number generator. In 2006 IEEE Symposium on Security and Privacy (S P'06). 15 pp.-385.
- Peter Guttman. 2004. Cryptographic Security Architecture: Design and Verification. Springer.
- K. Inayah, B. E. Sukmono, R. Purwoko, and S. Indarjani. 2013. Insertion attack effects on standard PRNGs ANSI X9.17 and ANSI X9.31 based on statistical dis- tance tests and entropy difference tests. In 2013 International Conference on Com- puter, Control, Informatics and Its Applications (IC3INA). 219-224.
- Gilles Pokam, Cristiano Pereira, Klaus Danne, Lynda Yang, Sam King, and Jesep Torrellas. 2009. Hardware And Software Approaches For Deterministic Multi- Processor Replay Of Concurrent Programs. Intel Technology Journal 13 (2009). Issue 4.
- Hany Ragab, Alyssa Milburn, Kaveh Razavi, Herbert Bos, and Cristiano Giuffrida. 2021. CrossTalk: Speculative Data Leaks Across Cores Are Real. In IEEE Security and Privacy. Paper=https:// download.vusec.net/ papers/ crosstalk_sp21.pdfWeb=https:// www.vusec.net/ projects/ crosstalkCode=https:// github.com/ vusec/ ridl
- J. K. Salmon, M. A. Moraes, R. O. Dror, and D. E. Shaw. 2011. Parallel random numbers: As easy as 1, 2, 3. In SC '11: Proceedings of 2011 International Conference for High Performance Computing, Networking, Storage and Analysis. 1-12.
- L. H. C. Tippett. 1927. Random Sampling Numbers: Tracts for Computers, No. XV. University Press, Cambridge.
- Jon von Neumann. 1951. Various Techniques Used in Conneciton With Random Digits. Journal of Research, Applied Math Series 3 (1951), 36- 38. https:// mcnp.lanl.gov/ pdf_files/ nbs_vonneumann.pdf Summary written by George E. Forsythe.
- V. M. Weaver, D. Terpstra, and S. Moore. 2013. Non-determinism and overcount on modern hardware performance counter implementations. In 2013 IEEE Inter- national Symposium on Performance Analysis of Systems and Software (ISPASS). 215-224.
- Michael Wertheimer. 2015. Encryption and the NSA Role in International Stan- dards. (2015). http:// www.ams.org/ notices/ 201502/ rnoti-p165.pdf Note: At the time of publication, Michael Wertheimer was the Director of Research at the US National Security Agency. Last Accessed May 30, 2020.
Cited by
Sorted by citations
Medical imaging deep learning with differential privacy
Ziller, A., Usynin, D., Braren, R., Makowski, M., Rueckert, D., & Kaissis, G. (2021) · Scientific Reports · doi:10.1038/s41598-021-93030-0 · Cited by 172 · see citing works
The 2020 Census Disclosure Avoidance System TopDown Algorithm
Abowd, J. M., Ashmead, R., Cumings-Menon, R., Garfinkel, S., Heineck, M., Heiss, C., Johns, R., Kifer, D., Leclerc, P., Machanavajjhala, A., et al. (2022) · Harvard Data Science Review · doi:10.1162/99608f92.529e3cb9 · Cited by 83 · see citing works
Self-citationDiscrete Gaussian for Differential Privacy
Canonne, C. L., Kamath, G., & Steinke, T. (2022) · Journal of Privacy and Confidentiality · doi:10.29012/jpc.784 · Cited by 73 · see citing works
Talwar, K. (2022) · DROPS (Schloss Dagstuhl – Leibniz Center for Informatics) · doi:10.4230/lipics.forc.2022.7 · Cited by 42 · see citing works
Differential Privacy and the 2020 US Census
Garfinkel, S. (2022) · doi:10.21428/2c646de5.7ec6ab93 · Cited by 21 · see citing works
Self-citationModel Integration in Computational Biology: The Role of Reproducibility, Credibility and Utility
Karr, J. R., Malik‐Sheriff, R. S., Osborne, J. M., González‐Parra, G., Forgoston, E., Bowness, R., Liu, Y., Thompson, R. N., Garira, W., Barhak, J., et al. (2022) · Frontiers in Systems Biology · doi:10.3389/fsysb.2022.822606 · Cited by 17 · see citing works
Secure Random Sampling in Differential Privacy
Holohan, N., & Braghin, S. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-88428-4_26 · Cited by 14 · see citing works
Differentiaalisesti yksityinen approksimatiivinen Bayes-päättely todennäköisyysmalleille
Jälkö, J. (2023) · Aaltodoc (Aalto University) · Cited by 14 · see citing works
Differentially Private Release of Israel's National Registry of Live Births
Hod, S., & Canetti, R. (2025) · IEEE Symposium on Security and Privacy (S&P) · doi:10.1109/sp61157.2025.00101 · Cited by 12 · see citing works
Privacy-preserving data sharing via probabilistic modeling
Jälkö, J., Lagerspetz, E., Haukka, J. K., Tarkoma, S., Honkela, A., & Samuel Kaski, A. (2021) · Patterns · doi:10.1016/j.patter.2021.100271 · Cited by 11 · see citing works
Overlook: Differentially Private Exploratory Visualization for Big Data
Budiu, M., Thaker, P., Gopalan, P., Wieder, U., & Zaharia, M. (2022) · Journal of Privacy and Confidentiality · doi:10.29012/jpc.779 · Cited by 8 · see citing works
Exploring the intersection of natural sciences and information technology via entropy and randomness
Luescher, A. M., Heckel, R., & Grass, R. N. (2025) · Nature Communications · doi:10.1038/s41467-025-62353-1 · Cited by 10 · see citing works
Madje, U. P., & Pande, M. B. (2024) · International Conference on Trends in Quantum Computing and Emerging Business Technologies (TQCEBT) · doi:10.1109/tqcebt59414.2024.10545237 · Cited by 5 · see citing works
Memristor-based security primitives
Gutiérrez, S. V. (2022) · IEEE European Test Symposium (ETS) · doi:10.1109/ets54262.2022.9810403 · Cited by 5 · see citing works
d3p - A Python Package for Differentially-Private Probabilistic Programming
Prediger, L., Loppi, N., Samuel Kaski, A., & Honkela, A. (2022) · Proceedings on Privacy Enhancing Technologies · doi:10.2478/popets-2022-0052 · Cited by 4 · see citing works
Applications of certified randomness
Amer, O., Chakrabarti, S., Chakraborty, K., Eloul, S., Kumar, N., Lim, C., Liu, M., Niroula, P., Satsangi, Y., Shaydulin, R., et al. (2025) · Nature Reviews Physics · doi:10.1038/s42254-025-00845-1 · Cited by 3 · see citing works
A Qualitative Analysis of Practical De-Identification Guides
Guo, W., Kishore, A., Aviv, A. J., & Mazurek, M. L. (2024) · ACM SIGSAC Conference on Computer and Communications Security (CCS) · doi:10.1145/3658644.3690270 · Cited by 2 · see citing works
Privacy-Aware Rejection SamplingPreprint
Awan, J., & Rao, V. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2108.00965 · Cited by 2 · see citing works
Cooperative Local Differential Privacy: Securing Time Series Data in Distributed Environments
Singh, B. C., Hossain, M. J., Dı́az, R., Roy, S., Mukkamala, R., & Shetty, S. S. (2025) · IEEE International Conference on Mobile Cloud Computing, Services, and Engineering (MobileCloud) · doi:10.1109/mobilecloud66020.2025.00007 · Cited by 1 · see citing works
Natural differential privacy—a perspective on protection guarantees
Altman, M., & Cohen, A. (2023) · PeerJ Computer Science · doi:10.7717/peerj-cs.1576 · Cited by 1 · see citing works
PatDiscover: Privacy-Preserving Discoverability of Patients
Pennekamp, J., Lohmöller, J., Pressel, N., Geisler, S., Mottaghy, F. M., & Wehrle, K. (2025) · Annual Computer Security Applications Conference Workshops (ACSAC Workshops) · doi:10.1109/acsacw69556.2025.00022 · Cited by 1 · see citing works
Kutschera, S., Slany, W., Ratschiller, P., Gursch, S., & Dagenborg, H. (2023) · Entropy · doi:10.3390/e25060854 · Cited by 1 · see citing works
Hoofnagle, C. J., & Garfinkel, S. (2021) · Cambridge University Press eBooks · doi:10.1017/9781108883719.019
Self-citationHoofnagle, C. J., & Garfinkel, S. (2021) · Cambridge University Press eBooks · doi:10.1017/9781108883719.016
Self-citationEnhancing Provable Security and Efficiency of Permutation-Based DRBGs
Chung, W., Hwang, S., Kim, H., & Lee, J. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-01901-1_15
Characterizing the System Overhead of Discrete Noise Generation for Differential Privacy
Kong, S., Kim, D., Seo, E., & Maeng, K. (2025) · IEEE Computer Architecture Letters · doi:10.1109/lca.2025.3627101
Quantum Key Distribution for Secure Financial Data Exchange
Wawage, P., Ikhar, S., Manhas, R., Sudarson, J., Rahmatullayev, B., & Maratuly, S. N. (2025) · International Conference on Mathematics, Actuarial Science, Computer Science and Statistics (MACS) · doi:10.1109/macs68476.2025.11453349
Adversarially Mixed Secret Key Generation for Side-Channel Defense for the Cloud
Karanam, V. S. S. L., Sakkaff, Z., & Balasooriyalage, P. (2026) · IEEE Secure Development Conference (SecDev) · doi:10.1145/3805773.3805999
- Authors
- Garfinkel, S. L., & Leclerc, P.
- Affiliations
- United States Census Bureau (US)
- Venue
- Proceedings of the 19th Workshop on Privacy in the Electronic Society
- Publisher
- ACM
- Published
- 2020-11-09
- Pages
- 73-86
- OpenAlex
- W3084157897
- Cited by
- 29
- References
- 23
- Citations by year
- 2026: 1 · 2025: 8 · 2024: 2 · 2023: 3 · 2022: 9 · 2021: 6
- FWCI
- 1.72
- Type
- conference paper
- Language
- en
- Open access
- green
- arxiv.org
- Licence
- Locations
- Workshop on Privacy in the Electronic Society (publishedVersion)
- arXiv (Cornell University) (submittedVersion, public-domain)
- arXiv (Cornell University) (public-domain)
- Funders
- US Census Bureau
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Cryptographic Implementations and Security
- Physical Sciences › Computer Science › Artificial Intelligence › Privacy-Preserving Technologies in Data
- Physical Sciences › Computer Science › Artificial Intelligence › Cryptography and Data Security
- Keywords
- differential privacy, random number generation, 2020 Census, Mersenne Twister, hardware random number generator, trusted hardware, privacy-preserving data collection
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
