Automating accountability? Privacy policies, data transparency, and the third party problem
Bibliographic record
Abstract
We have a data transparency problem. Currently, one of the main mechanisms we have to understand data flows is through the self-reporting that organizations provide through privacy policies. These suffer from many well-known problems, problems that are becoming more acute with the increasing complexity of the data ecosystem and the role of third parties – the affiliates, partners, processors, ad agencies, analytic services, and data brokers involved in the contemporary data practices of organizations. In this article, we argue that automating privacy policy analysis can improve the usability of privacy policies as a transparency mechanism. Our argument has five parts. First, we claim that we need to shift from thinking about privacy policies as a transparency mechanism that enhances consumer choice and see them as a transparency mechanism that enhances meaningful accountability. Second, we discuss a research tool that we prototyped, called AppTrans (for Application Transparency), which can detect inconsistencies between the declarations in a privacy policy and the actions the mobile application can potentially take if it is used. We used AppTrans to test seven hundred applications and found that 59.5 per cent were collecting data in ways that were not declared in their policies. The vast majority of the discrepancies were due to third party data collection such as advertising and analytics. Third, we outline the follow-on research we did to extend AppTrans to analyse the information sharing of mobile applications with third parties, with mixed results. Fourth, we situate our findings in relation to the third party issues that came to light in the recent Cambridge Analytica scandal and the calls from regulators for enhanced technical safeguards in managing these third party relationships. Fifth, we discuss some of the limitations of privacy policy automation as a strategy for enhanced data transparency and the policy implications of these limitations.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.002 | 0.001 |
| Scholarly communication | 0.000 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".