We are pleased to announce the new Ethical Data Initiative research strand, and the thematic topic of our November Townhall meeting: Recommender systems.

We all know the feeling: You open Netflix, and BAM, a movie suggestion appears. You scroll TikTok, and every video seems tailor-made to keep you interested. That’s a recommender system at work—the powerful engine that drives everything from your Amazon shopping cart to your Facebook feed. They feel helpful, but what if these invisible algorithms are doing more than just entertaining you? What if they’re quietly shaping our world, influencing our decisions, and even shifting our democracy?

Recommender systems are among the most influential yet least understood technologies of our time. They are designed to keep us engaged, but this single-minded goal has profound societal consequences, which are making headlines daily.

We have probably all heard of the Echo Chamber Effect: an environment or ecosystem in which participants encounter beliefs that amplify or reinforce their preexisting beliefs. By feeding us content similar to what we already like, recommenders create “filter bubbles.” This phenomenon has been heavily linked to political polarisation, where people are increasingly only exposed to views they already agree with, making compromise difficult and fueling extremism. Increasing the diversity of some recommendations is not enough to break this challenging status quo. 

There are real-world consequences to the use of recommender systems, extending beyond the realms of social media and e-commerce. These algorithms are not just used to serve us videos and products online. They also power platforms for reviews and funding, dating apps, and advanced bibliographical research tools. If the data they are trained on reflects historical prejudices—say, against certain demographics—the algorithm can learn and amplify that bias, unfairly locking people out of opportunities. This isn’t just a technical glitch; it’s an ethical crisis.

At the Ethical Data Initiative, we believe that the technology that connects us should also serve us fairly. That’s why the work of our two heads of research, Silvia Milano (TUM) and Stephan Guttinger (Exeter), is so critical to our mission.

Guttinger and Milano’s research focuses specifically on how recommenders operate in high-stakes environments, such as science and academia, where fairness, transparency, and the prevention of bias are paramount to the integrity and public trust in research. Their work investigates how these systems can be designed not just for efficiency, but for social good.

Milano, in her influential work on ethical considerations in AI, emphasised the need to move beyond simple risk management to ensure that computational systems, such as recommenders, embody fundamental human values (Milano, 2021). They are pushing for the next generation of algorithms: ones that break us out of unexamined and harmful habits, not lock us in.

Join the Ethical Data Initiative (EDI) Townhall

We are excited to announce a new Ethical Data Initiative research theme and the topic of our November Townhall: Recommender Systems.

Based on the research of our Heads of Research Silvia Milano (TUM) and Stephan Guttinger (University of Exeter)

Discover and discuss how these powerful recommendation engines truly work and what groundbreaking ethical research is being done to fix their most dangerous flaws. This is your chance to move beyond the headlines and understand the code that controls your daily life.

Why You Should Care: If you use a smartphone, a streaming service, or social media, this Townhall is about you. Don’t miss this opportunity to understand the mechanics of your digital world and empower yourself to demand fairer, more ethical technology.

Citation:  Milano, S., Taddeo, M. & Floridi, L. Recommender systems and their ethical challenges. AI & Soc 35, 957–967 (2020). https://doi-org.uoelibrary.idm.oclc.org/10.1007/s00146-020-00950-y