Categories
General

GDPR: How a lack of enforcement came to undermine its ambition and reputation

This article was written for the Inspired Research magazine of the Depatment of Computer Science of the University of Oxford. URL: https://www.cs.ox.ac.uk/innovation/inspiredresearch/InspiredResearch–summer23_FINAL_web.pdf

Tracking, the collection and sharing of behavioural data about individuals, is widely used by app developers to analyse and optimise apps and to show ads. It also is a significant and ubiquitous threat in mobile apps, and often violates data protection and privacy laws.

Previously, our research group led by Prof Sir Nigel Shadbolt analysed 1 million Android apps from the Google Play Store from 2017. We found that about 90% of those apps could share data with Alphabet (the parent company of Google), and 40% with Facebook (now renamed ‘Meta’). The data practices in children’s apps were particularly worrisome, which is why our research group – in response – established a dedicated research strand on Kids Online Anonymity & Lifelong Autonomy (KOALA), led by Dr Jun Zhao. Our findings led to major news coverage back then (including by the Financial Times). This underlines the extent to which those data practices violated individuals’ privacy expectations. Google even issued a public response to our findings, in which they tried to cast doubt over the validity of our (peer-reviewed) methodology.

Categories
General

Summer update

It’s been a bit quiet on here over recent months. A lot has happened. I finished my computer science PhD in Oxford, and moved to a new country to start a new job. Now, I am working at the Law & Tech Lab of the Law Faculty of Maastricht University, in the Netherlands. The change is obviously taking up a little bit of time.

Besides, I’m working on a new book project. This means that, over the last few weeks and months, I’ve nearly been writing a short blog article every day. Hence, I’ve not been so keen on writing even more on this blog. Feel free to reach out though if you want to learn more about my ongoing projects. Contact details are at the bottom of the homepage.

Categories
General

A Tracking-Free Mobile Ecosystem?

Over recent months, the pressure on providers of app tracking technology has been increasing. In February 2022, the Belgian data protection authority found that the IAB’s Transparency & Consent Framework is in violation of EU/UK data protection law. Among other aspects, the authority argued that the IAB is in fact a data controller and not just those organisations that use the IAB framework. This ruling by the authority represents one of the first applications of the EJC’s rulings on joint controllership within the context of web and mobile tracking. The ruling underlines that those who design the technical infrastructure behind the tracking ecosystem bear responsibility for their design decisions under EU/UK data protection law. Is this the beginning of the end of surveillance capitalism?

Categories
General

Research at the Gap?

Looking around in my PhD, many of my peers seem to have an obsession with one key thing: identifying the right research gap. The promise is that, by finding your gap in the existing literature, you can then conduct the thus far missing research and make important contributions with your PhD.

As a result of that thinking, many students spend a significant amount of their time studying the existing research landscape meticulously. The wish to get a grasp of the state-of-the-art is understandable. After all, there’s a certain bar to pass for academic work and it’s difficult to know that bar right at the start of your PhD.

I do strongly think that the approach of searching for research gaps and trying to fill them is a relatively safe and tested approach to succeed in a PhD. This is because many academic venues — journals and conferences — adopt a similar thinking when assessing the contribution of your research: if your own research complements and sufficiently exceeds the status quo, then your work is more likely to get published.

However, the focus on ‘research gaps’ evokes feelings in me like when I see my dentist. He, too, makes a meticulous effort to find gaps and fill them, within the cosmos of my dental hygiene. This work is utterly important. It will, however, not change the world and lead to great leaps in society. Neither is an overly focus on filling research gaps going to lead to great academic leaps. Rather, this creates iterative work instead of groundbreaking innovations.

The issue of a loss of innovation in science was recently covered in a widely discussed Nature paper. This paper seems to confirm my own observations around a mismatch of incentives between the academic publication process and the creation of groundbreaking scientific breakthroughs.

Hence my advice for young PhD students: If you want to create the most good for society, then better forget about the research gap, study widely and outside your immediate bubble, and aim for research leaps instead!

Categories
General

Federated analytics for public good

About contact-tracing apps in Covid-19 containment

For the blog of the Open Data Institute (ODI), I wrote a blog article that talks about what federated analytics is, how it relates to federated learning and how the approach was used in Covid-19 contact tracing. This post is related to ongoing research at the ODI on the use of federated learning for public, educational and charitable aims.

Categories
General

Ethics is all we need?

The computer science discipline undoubtedly has enormous impacts on our day-to-day lives. Increasingly, key parts of our society are run by algorithms. This has created a range of notable scandals, for example, when machine learning systems fail to account for the diversity of the human population. This is why many individuals, inside and outside of computer science, argue for compulsory ethics modules in the computer science curriculum. This would be similar to compulsory ethics education that is commonly part of business and finance degrees. I thus argue this thinking is too simple, and fails to identify the actual root problems.