Evaluating the Contextual Integrity of False Positives in Algorithmic Travel Surveillance

Alan Medlar; Alina Wernick; Dorota G{\l}owacka; Sofia S\"oderholm

REVIEW 3 major objections 4 minor 84 references

Evaluating the Contextual Integrity of False Positives in Algorithmic Travel Surveillance

T0 review · 3 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A nationally representative survey finds Finns accept double-digit false-positive rates in algorithmic travel pre-screening, with air travel more accepted than sea travel.

desk verdict New Finnish survey data on public tolerance for false positives in travel surveillance is genuinely useful, though the headline false positive rates are model-dependent and the abstract misreports gender-specific numbers as overall rates. read the letter →

arxiv 2506.00218 v1 pith:FRA4DK2S submitted 2025-05-30 cs.CY

Alina Wernick , Alan Medlar , Sofia Söderholm , Dorota G{l}owacka This is my paper

classification cs.CY

keywords privacytrade-offsurveillancecontextualintegrityPNRDirectivefalsepositivesdataprotectionpublicsecurityalgorithmic

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper asks how much false-positive error the public will tolerate in algorithmic passenger pre-screening, a statistical feature of the EU's passenger-name-record (PNR) surveillance system that is hidden from passengers by law. Using a nationally representative survey of 1,550 Finnish adults, it finds that respondents judged double-digit false-positive rates legitimate under a contextual-integrity standard, roughly 11-13% combined for sea and air travel, and between 9.7% and 14.7% depending on travel mode and gender. It also finds air travel surveillance consistently more acceptable than sea travel, and women accept more false positives than men. The authors argue this bottom-up measure of legitimacy complements top-down rights-based EU regulation, which has struggled with the base-rate fallacy. If the finding holds, critiques of mass surveillance that focus on individual rights violations may not match what the public actually accepts.

What carries the argument

The central device is a vignette-based trade-off task paired with interval regression. Respondents see 200 flagged passengers with a random number of true and false positives, choose between prioritizing security or protecting the privacy of innocent passengers, and that binary choice is converted into an interval on the acceptable number of false positives, for example choosing security at 10 false positives means the acceptable count lies in the interval (10, 200]. Assuming the acceptable count follows a log-normal distribution, the paper fits interval regression models to estimate the mean acceptable false-positive count, which it then reports as a false-positive rate. The collective-criticism origin of the method allows nearby vignette values to inform the same estimate.

What would settle it

Run the same vignette with a different denominator, for example 500 or 1,000 flagged passengers, and also ask respondents directly for the maximum acceptable number of false alarms; if the estimate does not scale with the denominator, or direct answers diverge from the regression-based threshold, the reported rates are an artifact of the 200-passenger frame and the assumed curve.

Watch

Extended reading notes

Core claim

Working within the contextual-integrity framework, the paper claims that the Finnish public's perception of when algorithmic travel surveillance becomes illegitimate can be quantified as a false-positive threshold, and that this threshold is much higher than rights-based critiques would predict. In a vignette where 200 passengers are flagged and a randomly sampled number turn out to be false alarms, respondents chose between prioritising security and protecting innocent passengers' privacy; interval regression on those binary choices yields estimates of 22.10 acceptable false positives out of 200 (11.1%) for sea travel and 25.83 (12.9%) for air travel when pooling genders, with the gender-specific figures spanning 9.7% to 14.7%. The estimates are significantly higher for air than for sea travel, and women's thresholds are about 29% higher than men's. The authors take this as evidence that the contextual integrity of the information flow is not breached by high false-positive counts, and that acceptability tracks the travel context rather than the identical rights violations suffered by the flagged passengers.

Load-bearing premise

The load-bearing assumption is that a person who says security comes first would accept every number of false alarms from the one shown up to 200, and that the statistical curve used to turn those either-or answers into exact thresholds is shaped correctly; if people would accept more than 200, or the curve is misshapen, the reported percentages change.

Editorial extensions

If this is right

At the 12.7% air-travel false-positive rate the authors estimate, a fully operational EU-wide PNR system would require on the order of 140,000 manual checks per day, raising the question of whether the accepted error rates are operationally sustainable.
Because respondents found air-travel surveillance more acceptable than sea-travel surveillance even though the privacy consequences are identical, expanding algorithmic screening to new transport modes is likely to face more public resistance than continuing it where it already exists.
The public's acceptance of high false-positive counts suggests that legal challenges to mass surveillance that argue from individual rights may miss the statistical, systemic character of the harm; legitimacy arguments would need to engage with aggregate error rates rather than only individual redress.
Respondents' thresholds did not vary with age or with the number of true positives shown, so within the tested range the acceptability of false positives appears stable across those dimensions.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

Going beyond the paper, the same vignette run in a lower-trust or more ethnically diverse population would test whether the double-digit thresholds are a Finnish high-trust result; the paper itself flags this cultural-context caveat.
Because the vignette fixes the flagged pool at 200 passengers, the estimates may be anchored to that denominator; varying the pool size in a replication would show whether the 10-15% rates are a framing artifact.
If many respondents who chose security would have accepted more than 200 false alarms, the reported rates are lower bounds, and true public tolerance could be even higher than the abstract's headline figures.
The interval-regression design could be combined with a direct elicitation question, such as asking respondents how many false alarms are too many, to test whether respondents hold a stable internal threshold or merely react to the presented trade-off.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The paper is worth reading for one reason: it gives the first nationally representative estimate of how much false positive error the public will tolerate in EU PNR-style travel surveillance. The Finnish data (N=1550) and the air-versus-sea comparison are new, and the qualitative finding — Finns tolerate double-digit false positive rates, and tolerate them more for air than sea travel — is reasonably well supported. The method is a novel application of the authors' collective criticism technique to contextual integrity, which is a genuine empirical contribution even if the underlying method is not new. Credit where due: the survey design is careful, the limitations section is honest, and the attitudinal questions corroborate the vignette results.

The soft spots are real, though not fatal. The precise headline rates (9.7% and 14.7% in the abstract) are not direct measurements. They come out of an interval regression that converts each binary vignette response into an interval capped at the vignette's 200 passengers, then fits a log-normal distribution to those intervals. The 200 cap and the log-normal assumption are both arbitrary and untested, and the paper reports no robustness checks. A respondent who prioritizes security at 30 false positives is treated as accepting everything up to 200 — that's a modeling choice, not a fact. And the abstract is worse than imprecise: it presents the lowest and highest gender-specific rates (male/sea and female/air) as if they were the overall sea and air estimates. Table 2's actual overall rates are 11.1% and 12.9%. That should be corrected.

That said, the stress-test's claim that the rates are "artifacts" goes too far. The raw response pattern — a majority choosing security even at high false positive counts — is directly observable, and the significance tests for travel mode and gender are standard and fine. The qualitative conclusion that Finns are highly tolerant of false positives, and more tolerant in air travel, does not depend on the log-normal assumption. What depends on the assumption is the exact number, and that number is being overstated in the abstract.

The paper would be stronger with a sensitivity analysis (different caps, maybe a nonparametric description of the response curve) and with data/code released. As it stands, it is a solid empirical contribution for people working on surveillance, contextual integrity, and EU data protection law. It deserves a serious referee, and I'd want to see it after revision. I would cite it for the new data.

Referee Report

3 major / 4 minor

Summary. The paper reports a nationally representative Finnish survey (N=1550) with randomized air-travel and sea-travel conditions, using a vignette that presents a binary choice between prioritizing security and prioritizing privacy at a given number of false positives. Responses are converted into intervals and analyzed with log-normal interval regression, from which the authors estimate 'legitimate' false-positive counts and rates. The paper claims that Finns accept very high false-positive rates (9.7% for sea and 14.7% for air in the abstract), that air travel is more acceptable than sea travel, and that these findings challenge purely rights-based justifications of algorithmic surveillance. The discussion connects the results to contextual integrity, the base-rate fallacy, and the limits of the EU legal framework.

Significance. If the quantitative estimates were credible, the paper would make a distinctive empirical contribution to the surveillance-privacy literature: it offers one of the first population-level measurements of acceptable false-positive rates for PNR-style pre-screening, and it explicitly frames the result as a challenge to individual-rights-based legitimacy arguments. The study has clear strengths: a large, stratified sample; randomized experimental conditions; a transparent questionnaire appendix in both English and Finnish; consistent between-condition differences in the attitudinal items; and an unusually honest limitations section that acknowledges normalcy bias and the human-in-the-loop simplification. The paper is also careful to distinguish the descriptive use of contextual integrity from a full normative evaluation. However, the headline false-positive rates are not directly elicited quantities; they are outputs of an interval regression whose two central assumptions—the 200-passenger ceiling and the log-normal latent distribution—are neither justified nor stress-tested.

major comments (3)

[§4.3.3, Table 2] The interval transformation in Section 4.3.3 is the sole source of the reported point estimates. A security-prioritizing response at F false positives is converted to the interval (F, 200], and a privacy response to [0, F), which makes the vignette's 200-passenger denominator an upper bound. The log-normal distribution is assumed without diagnostics, and the vignette only showed F values from 1 to 30 (§4.1.3), so the model's upper confidence limits (e.g., 33.03 in Table 2) and any estimate treated as a population threshold depend on extrapolation beyond the displayed range. The Table 2 rates are the exponentiated linear predictor—the median of the assumed log-normal distribution—divided by 200, not directly measured acceptability rates. Please report robustness checks (for example, alternative distributional assumptions, nonparametric interval bounds, a sensitivity analysis varying the vignette denominator, or at minimum the raw proportion choosing security at each F by condition) before these numbers are presented as legitimate false-positive rates.
[Abstract, §5.3, §6] The abstract's headline rates '9.7% and 14.7% for sea and air travel' do not match Table 2's combined estimates of 11.1% for sea and 12.9% for air; 9.7% and 14.7% are instead the gender-specific extremes (male/sea and female/air). The Discussion also refers to an air-travel rate of 12.7%, which differs from the Table 2 air-travel combined value of 12.9%. The paper should consistently report a single set of estimates, clearly distinguishing pooled estimates from gender-specific subgroups.
[§5.2] The ANOVA tests used for model selection are nested comparisons of interval-regression models, so the conclusions that age and the number of true positives do not affect acceptance are conditional on the same interval/log-normal specification that produces the point estimates. Reporting raw associations—for example, the proportion of security-prioritizing responses by true-positive count and by age group—would provide a specification-free check of these null results and would strengthen the model-selection claim.

minor comments (4)

[§3.2] 'Course-grained' should be 'coarse-grained'.
[§6] 'Rights-bases approaches' should be 'rights-based approaches'.
[References] Reference [69] spells the author's name as 'Chrstian Thönnes'; the correct spelling is 'Christian Thönnes'.
[References] Reference [23] is dated 'Retrieved 21 December 2025', which postdates the arXiv submission date of the manuscript; please verify the retrieval date.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the false-positive thresholds are model outputs from new survey data, and the self-cited collective criticism method is external, not result-forcing.

full rationale

The paper's derivation chain runs from new, randomly assigned vignette responses to intervals, then to interval-regression estimates, then to false-positive rates. In Section 4.3.3, a security-prioritizing response at F false positives is coded as (F, 200] and a privacy-prioritizing response as [0, F); these intervals are the dependent variable in a log-normal interval regression, and the reported counts in Table 2 are exp(linear predictor). The abstract's 9.7% and 14.7% rates are these model outputs divided by the fixed vignette denominator of 200. Nothing in this chain defines the target conclusion in terms of itself: the 'legitimate false positive count' is not an input or a fitted parameter renamed as a prediction, but a statistical summary of newly collected survey responses. The only self-citation is to Medlar et al. [43] (and [44]) for the collective criticism interval transformation. That citation supplies the elicitation method, not the empirical conclusion, and it is not a uniqueness theorem or an ansatz whose content is the paper's own result. The log-normal distributional assumption and the 200-passenger ceiling are real identification and robustness concerns, and the abstract's headline rates are actually gender-specific extremes from Table 2 rather than overall estimates, but these are validity/correctness issues, not circularity. The paper also acknowledges its main modeling simplifications in Section 4.4. The central empirical claim is therefore self-contained with respect to the survey data, and the self-citation is not load-bearing in the circularity sense.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central estimates rest on three hand-chosen design parameters: the 200-passenger vignette cap, the log-normal distribution assumption, and the shown ranges of false positives. These are not fitted to data but are load-bearing choices that determine the reported thresholds. No new entities are introduced.

free parameters (3)

Total passengers in vignette = 200
Chosen by hand as the denominator for false positive rates and as the upper bound for the interval when a respondent prioritizes security. This arbitrary cap directly affects the point estimates.
Distributional assumption for acceptable false positives = log-normal
Assumed for the interval regression dependent variable, stated in Section 4.3.3. Affects the mean estimates and confidence intervals.
Shown ranges in vignettes = true positives 0-4, false positives 1-30
Selected by the authors with the stated reasoning that true positives are rare and the acceptable false positive range is unknown. This limits what can be estimated beyond 30 false positives.

assumptions (4)

domain assumption Collective criticism interval mapping correctly represents privacy-security trade-offs
The paper assumes that a binary choice between security and privacy reveals an interval of acceptable false positive counts, following Medlar et al. (2022). This is a behavioral assumption not independently validated here.
standard math Interval regression with censoring is an appropriate estimator
The analysis uses interval regression on transformed responses, a standard statistical method, but its suitability for this type of vignette data is asserted rather than tested.
domain assumption Sample balancing makes the sample representative
The paper relies on sample balancing to reduce non-response bias, assuming the post-stratified weights recover population estimates. This is a common but unverifiable assumption.
domain assumption Contextual integrity is a valid descriptive lens for legitimacy
The paper uses Nissenbaum's contextual integrity framework as a descriptive complement to rights-based approaches, a theoretical choice that shapes interpretation.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Evaluating the Contextual Integrity of False Positives in Algorithmic Travel Surveillance." pith.science (2026). https://pith.science/paper/FRA4DK2S

@misc{pith2026250600218,
  author       = {Pith},
  title        = {Pith review of: Evaluating the Contextual Integrity of False Positives in Algorithmic Travel Surveillance},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FRA4DK2S}},
  note         = {Machine review of arXiv:2506.00218}
}

read the original abstract

International air travel is highly surveilled. While surveillance is deemed necessary for law enforcement to prevent and detect terrorism and other serious crimes, even the most accurate algorithmic mass surveillance systems produce high numbers of false positives. Despite the potential impact of false positives on the fundamental rights of millions of passengers, algorithmic travel surveillance is lawful in the EU. However, as the system's processing practices and accuracy are kept secret by law, it is unknown to what degree passengers are accepting of the system's interference with their rights to privacy and data protection. We conducted a nationally representative survey of the adult population of Finland (N=1550) to assess their attitudes towards algorithmic mass surveillance in air travel and its potential expansion to other travel contexts. Furthermore, we developed a novel approach for estimating the threshold, beyond which, the number of false positives breaches individuals' perception of contextual integrity. Surprisingly, when faced with a trade-off between privacy and security, even very high false positive counts were perceived as legitimate. This result could be attributed to Finland's high-trust cultural context, but also raises questions about people's capacity to account for privacy harms that happen to other people. We conclude by discussing how legal and ethical approaches to legitimising algorithmic surveillance based on individual rights may overlook the statistical or systemic properties of mass surveillance.

Figures

Figures reproduced from arXiv: 2506.00218 by the authors.

**Figure 2.** Example passport photos used in the false positives [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

84 extracted references · 77 canonical work pages

[1]

Noura Abdi, Xiao Zhan, Kopo M Ramokapane, and Jose Such. 2021. Privacy norms for smart home personal assistants. In Proceedings of the 2021 CHI conference on human factors in computing systems . 1–14

2021
[2]

Noah Apthorpe, Yan Shvartzshnaider, Arunesh Mathur, Dillon Reisman, and Nick Feamster. 2018. Discovering smart home internet of things privacy norms using contextual integrity. Proceedings of the ACM on interactive, mobile, wearable and ubiquitous technologies 2, 2 (2018), 1–23

2018
[3]

Noah Apthorpe, Sarah Varghese, and Nick Feamster. 2019. Evaluating the Con- textual Integrity of Privacy Regulation: Parents’ IoT Toy Privacy Norms Versus COPPA. In 28th USENIX security symposium (USENIX security 19) . 123–140

2019
[4]

Maya Bar-Hillel. 1980. The base-rate fallacy in probability judgments. Acta Psychologica 44, 3 (1980), 211–233

1980
[5]

Adam Barth, Anupam Datta, John C Mitchell, and Helen Nissenbaum. 2006. Privacy and contextual integrity: Framework and applications. In 2006 IEEE symposium on security and privacy . IEEE

work page 2006
[6]

Sebastian Benthall and Rachel Cummings. 2024. Integrating differential privacy and contextual integrity. In Proceedings of the Symposium on Computer Science and Law. 9–15

work page 2024
[7]

Sebastian Benthall, Seda Gürses, Helen Nissenbaum, et al . 2017. Contextual integrity through the lens of computer science. Foundations and Trends ® in Privacy and Security 2, 1 (2017), 1–69

work page 2017
[8]

Sebastian Benthall and Ido Sivan-Sevilla. [n. d.]. Regulatory CI: Adaptively Regulating Privacy as Contextual Integrity

work page

Show all 84 references

[9]

Michael Birnhack. 2011. A Quest for a Theory of Privacy: Context and Con- trol’(2011). Jurimetrics 51 (2011), 447

2011
[10]

Anu Bradford. 2023. Digital empires: The global battle to regulate technology . Oxford University Press

2023
[11]

Evelien Brouwer, Elspeth Guild, Stefan Salomon, and Christian Thönnes. 2023. The Future of the European Security Architecture-a Debate Series. Originally published on: https://verfassungsblog. de/category/debates/pnr-debate-series/, Max Planck Institute for the Study of Crime,...

2023
[12]

Matthew Browning and Bruce Arrigo. 2021. Stop and risk: Policing, data, and the digital age of discrimination. American Journal of Criminal Justice 46, 2 (2021), 298–316

2021
[13]

CJEU. 2022. Case C-817/19, Ligue des droits humains

2022
[14]

European Commission. 2022. Proposal for a REGULATION OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL laying down rules to prevent and combat child sexual abuse. https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/ ?uri=CELEX:52022PC0209

2022
[15]

Council of European Union. 2000. Charter of Fundamental Rights of European Union. https://www.europarl.europa.eu/charter/pdf/text_en.pdf

2000
[16]

Council of European Union. 2004. Council Directive 2004/82/EC of 29 April 2004 on the obligation of carriers to communicate passenger data. http://data.europa. eu/eli/dir/2004/82/oj

2004
[17]

Council of European Union. 2016. General Data Protection Regulation, Regulation (EU) 2016/679. http://data.europa.eu/eli/reg/2016/679/oj

2016
[18]

Council of European Union. 2016. Law Enforcement Directive, Directive (EU) 2016/2016/680. http://data.europa.eu/eli/dir/2016/680/oj

2016
[19]

Council of European Union. 2016. Passenger Name Record Directive, Directive (EU) 2016/681. https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri= CELEX:32016L0681

2016
[20]

Council of European Union. 2024. Artificial Intelligence Act, Regulation (EU) 2024/1689. http://data.europa.eu/eli/reg/2024/1689/oj

2024
[21]

John A Darling, Christopher L Jerde, and Adam J Sepulveda. 2021. What do you mean by false positive? Environmental DNA 3, 5 (2021), 879–883

2021
[22]

Giovanni De Gregorio and Pietro Dunn. 2022. The European risk-based ap- proaches: Connecting constitutional dots in the digital age. Common Market Law Review 59, 2 (2022)

2022
[23]

EDri. 2019. Why EU passenger surveillance fails its purpose. Retrieved 21 December 2025 from https://edri.org/our-work/why-eu-passenger-surveillance-fails-its- purpose/

2019
[24]

Finnair. 2024. Finnair privacy policy. Retrieved 19 November 2024 from https: //www.finnair.com/de-en/info/finnair-privacy-policy

2024
[25]

Alisa Frik, Julia Bernd, and Serge Egelman. 2023. A Model of Contextual Fac- tors Affecting Older Adults’ Information-Sharing Decisions in the US. ACM Transactions on Computer-Human Interaction 30, 1 (2023), 1–48

2023
[26]

Raphael Gellert. 2016. We have always managed risks in data protection law: understanding the similarities and differences between the rights-based and the risk-based approaches to data protection. Eur. Data Prot. L. Rev. 2 (2016), 481

2016
[27]

Audrey Guinchard. 2018. Taking proportionality seriously: The use of contextual integrity for a more informed and transparent analysis in EU data protection law. European Law Journal 24, 6 (2018), 434–457

2018
[28]

Kevin D Haggerty and Richard V Ericson. 2017. The surveillant assemblage. Surveillance, crime and social control (2017), 61–78

2017
[29]

Lucas Michael Haitsma. 2023. Regulating algorithmic discrimination through adjudication: the Court of Justice of the European Union on discrimination in algorithmic profiling based on PNR data. Frontiers in Political Science 5 (2023), 1232601

2023
[30]

David R Heise. 2010. Surveying cultures: Discovering shared conceptions and sentiments. John Wiley & Sons

2010
[31]

Michael Herrmann, Mireille Hildebrandt, Laura Tielemans, and Claudia Diaz
[32]

Mireille Hildebrandt. 2014. Location Data, Purpose Binding and Contextual Integrity: What’s the Message? Protection of Information and the Right to Privacy- A New Equilibrium? (2014), 31–62

2014
[33]

Transparency International. 2023. 2023 Corruption Perceptions Index - Explore Finland’s results — transparency.org. https://www.transparency.org/en/cpi/2023/ index/fin

2023
[34]

Brian Jordan Jefferson. 2018. Predictable policing: Predictive crime mapping and geographies of policing and race. Annals of the American Association of Geographers 108, 1 (2018), 1–16

2018
[35]

Rob Kitchin and Martin Dodge. 2014. Code/space: Software and everyday life. MIT Press

2014
[36]

Bert-Jaap Koops. 2021. The concept of function creep. Law, Innovation and Technology 13, 1 (2021), 29–56

2021
[37]

Douwe Korff. 2021. Opinion on the Broader and Core Issues Arising in the PNR Case Currently Before the CJEU (Case C-817/19). A vailable at SSRN 4436951 (2021)

2021
[38]

Douwe Korff. 2023. Did the PNR judgment address the core issues raised by mass surveillance? European Law Journal 29, 1-2 (2023), 223–236

2023
[39]

Douwe Korff and M Georges. 2015. Passenger Name Records, data mining & data protection: the need for strong safeguards. Council of Europe, Directorate General Human Rights and Rule of Law, T-PD 11 (2015)

2015
[40]

Kristian Lum and William Isaac. 2016. To predict and serve? Significance 13, 5 (2016), 14–19

2016
[41]

Thomas Lumley. 2004. Analysis of complex survey samples. Journal of statistical software 9 (2004), 1–19

2004
[42]

Thomas Lumley and Alastair J Scott. 2013. Two-sample rank tests under complex sampling. Biometrika 100, 4 (2013), 831–842

2013
[43]

Alan Medlar, Jing Li, Yang Liu, and Dorota Glowacka. 2022. Critiquing-based Modeling of Subjective Preferences. In Proceedings of the 30th ACM Conference on User Modeling, Adaptation and Personalization . 234–242

2022
[44]

Alan Medlar, Joel Pyykkö, and Dorota Glowacka. 2017. Towards Fine-Grained Adaptation of Exploration/Exploitation in Information Retrieval. In Proceedings of the 22nd International Conference on Intelligent User Interfaces . 623–627

2017
[45]

Timme Bisgaard Munk. 2017. 100,000 false positives for every real terrorist: Why anti-terror algorithms don’t work. First Monday (2017)

2017
[46]

Helen Nissenbaum. 2009. Privacy in Context: Technology, Policy, and the Integrity of Social Life. Stanford University Press

2009
[47]

Helen Nissenbaum. 2019. Contextual integrity up and down the data food chain. Theoretical inquiries in law 20, 1 (2019), 221–256

2019
[48]

Henrik Nolte, Miriam Rateike, and Michèle Finck. 2024. Robustness and Cyberse- curity in the EU Artificial Intelligence Act. InGenerative AI and Law (GenLaw ’24) Workshop at 41 International Conference on Machine Learning, Vienna, Austria

2024
[49]

OECD. 2024. OECD Survey on Drivers of Trust in Public Institutions 2024 Results - Country Notes: Finland. https://www.oecd.org/en/publications/oecd- survey-on-drivers-of-trust-in-public-institutions-2024-results-country- notes_a8004759-en/finland_596ba5da-en.html. Evaluating t...

2024
[50]

Henrik Palmer Olsen and Cornelius Wiesener. 2021. Beyond data protection concerns–the European passenger name record system. Law, Innovation and Technology 13, 2 (2021), 398–421

2021
[51]

The Finnish Act on the Openness of Government Activities. 1999. https: //www.finlex.fi/fi/laki/ajantasa/1999/19990621

1999
[52]

Tessa Oomen, João Gonçalves, and Anouk Mols. 2024. Rethinking Artificial Intelligence: Algorithmic Bias and Ethical Issues| Rage Against the Artificial Intel- ligence? Understanding Contextuality of Algorithm Aversion and Appreciation. International Journal of Communication 18...

2024
[53]

Elisa Orrù. 2022. The European PNR Directive as an instance of pre-emptive, risk-based algorithmic security and its implications for the regulatory framework

2022
[54]

Elizabeth O’Neill. 2022. Contextual integrity as a general conceptual tool for evaluating technological change. Philosophy & Technology 35, 3 (2022), 79

2022
[55]

Information Polity 27, 2 (2022), 131–146

2022
[56]

Francesca Palmiotto. 2024. When Is a Decision Automated? A Taxonomy for a Fundamental Rights Analysis. German Law Journal 25, 2 (2024), 210–236

2024
[57]

Elizabeth O’Neill. 2023. Balancing Caution and the Need for Change: The General Contextual Integrity Approach. Philosophy & Technology 36, 4 (2023), 68

2023
[58]

Aditya Ramesh, Prafulla Dhariwal, Alex Nichol, Casey Chu, and Mark Chen
[59]

Alexandra Prégent. 2025. Why you Should not use CI to Evaluate Socially Disruptive Technology. Philosophy & Technology 38, 1 (2025), 1–19

2025
[60]

Madelyn Sanfilippo, Brett Frischmann, and Katherine Standburg. 2018. Privacy as commons: Case evaluation through the governing knowledge commons frame- work. Journal of Information Policy 8 (2018), 116–166

2018
[61]

Madelyn Rose Sanfilippo and Brett Frischmann. 2023. Slow-governance in smart cities: An empirical study of smart intersection implementation in four US college towns. Internet Policy Review 12, 1 (2023)

2023
[62]

James B Rule. 2019. Contextual integrity and its discontents: A critique of Helen Nissenbaum’s normative arguments. Policy & Internet 11, 3 (2019), 260–279

2019
[63]

Andrew D Selbst. 2013. Contextual expectations of privacy. Cardozo L. Rev. 35 (2013), 643

2013
[64]

Yan Shvartzshnaider, Noah Apthorpe, Nick Feamster, and Helen Nissenbaum
[65]

Madelyn R Sanfilippo, Yan Shvartzshnaider, Irwin Reyes, Helen Nissenbaum, and Serge Egelman. 2020. Disaster privacy/privacy disaster. Journal of the Association for Information Science and Technology 71, 9 (2020), 1002–1014

2020
[66]

Nathalie A Smuha. 2021. Beyond the individual: governing AI’s societal harm. Internet Policy Review 10, 3 (2021)

2021
[67]

Christian Thönnes. 2022. A cautious green light for technology-driven mass surveillance. Verfassungsblog On Matters Constitutional (2022)

2022
[68]

Christian Thönnes. 2022. A directive altered beyond recognition: on the Court of Justice of the European Union’s PNR decision (C-817/19). Verfassungsblog On Matters Constitutional (2022)

2022
[69]

Giovanni Sileno, Alexander Boer, Geoff Gordon, and Bernhard Rieder. 2021. Like circles in the water: Responsibility as a system-level function. In International Workshop on AI Approaches to the Complexity of Legal Systems. Springer, 198–211

2021
[70]

Fair Trials. 2021. Automating injustice: The use of artificial intelligence & auto- mated decision-making systems in criminal justice in Europe

2021
[71]

Maria Tzanou. 2010. The EU as an emerging’Surveillance Society’: The function creep case study and challenges to privacy and data protection. ICL Journal 4, 3 (2010), 407–427

2010
[72]

Lena Ulbricht. 2018. When big data meet securitization. Algorithmic regulation with passenger name records. European Journal for Security Research 3, 2 (2018), 139–161

2018
[73]

Chrstian Thönnes and Niovi Vavoula. 2023. Automated predictive threat detection. Max Planck Institute for the Study of Crime, Security and Law Working Paper, Chapter 1

2023
[74]

Alina Wernick and Anna Artyushina. 2023. Future-proofing the city: A hu- man rights-based approach to governing algorithmic, biometric and smart city technologies. Internet policy review 12, 1 (2023), 1–26

2023
[75]

Patrick Williams and Eric Kind. 2019. Data-driven Policing: The hardwiring of discriminatory policing practices across Europe

2019
[76]

David Murakami Wood, Kirstie Ball, David Lyon, Clive Norris, and Charles Raab
[77]

Ibo van de Poel. 2022. Socially disruptive technologies, contextual integrity, and conservatism about moral change. Philosophy & technology 35, 3 (2022), 82

2022
[78]

Shikun Zhang, Yan Shvartzshnaider, Yuanyuan Feng, Helen Nissenbaum, and Norman Sadeh. 2022. Stop the spread: A contextual integrity perspective on the appropriateness of covid-19 vaccination certificates. In Proceedings of the 2022 ACM Conference on Fairness, Accountability, a...

2022
[79]

Michael Zimmer. 2018. Addressing conceptual gaps in big data research ethics: An application of contextual integrity. Social Media+ Society 4, 2 (2018). FAccT ’25, June 23–26, 2025, Athens, Greece Alina Wernick, Alan Medlar, Sofia Söderholm, and Dorota Głowacka A Regression Mo...

2018
[82]

Karen Yeung. 2019. Responsibility and AI. https://rm.coe.int/responsability- and-ai-en/168097d9c5

2019
[2006]

Surveillance Studies Network, UK (2006), 1–98

A report on the surveillance society. Surveillance Studies Network, UK (2006), 1–98

2006
[2016]

Privacy in location-based services: An interdisciplinary approach.SCRIPTed 13 (2016), 144

2016
[2019]

In Proceedings of the AAAI Conference on Human Computation and Crowdsourcing, Vol

Going against the (appropriate) flow: A contextual integrity approach to privacy policy analysis. In Proceedings of the AAAI Conference on Human Computation and Crowdsourcing, Vol. 7. 162–170
[2022]

arXiv preprint arXiv:2204.06125 (2022)

Hierarchical text-conditional image generation with clip latents. arXiv preprint arXiv:2204.06125 (2022)

2022 arXiv

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [1]

Noura Abdi, Xiao Zhan, Kopo M Ramokapane, and Jose Such. 2021. Privacy norms for smart home personal assistants. In Proceedings of the 2021 CHI conference on human factors in computing systems . 1–14

2021

[2] [2]

Noah Apthorpe, Yan Shvartzshnaider, Arunesh Mathur, Dillon Reisman, and Nick Feamster. 2018. Discovering smart home internet of things privacy norms using contextual integrity. Proceedings of the ACM on interactive, mobile, wearable and ubiquitous technologies 2, 2 (2018), 1–23

2018

[3] [3]

Noah Apthorpe, Sarah Varghese, and Nick Feamster. 2019. Evaluating the Con- textual Integrity of Privacy Regulation: Parents’ IoT Toy Privacy Norms Versus COPPA. In 28th USENIX security symposium (USENIX security 19) . 123–140

2019

[4] [4]

Maya Bar-Hillel. 1980. The base-rate fallacy in probability judgments. Acta Psychologica 44, 3 (1980), 211–233

1980

[5] [5]

Adam Barth, Anupam Datta, John C Mitchell, and Helen Nissenbaum. 2006. Privacy and contextual integrity: Framework and applications. In 2006 IEEE symposium on security and privacy . IEEE

work page 2006

[6] [6]

Sebastian Benthall and Rachel Cummings. 2024. Integrating differential privacy and contextual integrity. In Proceedings of the Symposium on Computer Science and Law. 9–15

work page 2024

[7] [7]

Sebastian Benthall, Seda Gürses, Helen Nissenbaum, et al . 2017. Contextual integrity through the lens of computer science. Foundations and Trends ® in Privacy and Security 2, 1 (2017), 1–69

work page 2017

[8] [8]

Sebastian Benthall and Ido Sivan-Sevilla. [n. d.]. Regulatory CI: Adaptively Regulating Privacy as Contextual Integrity

work page