Public Values in the Age of Personalised News 

A joint research project by the Civic Machines Lab, Queensland University of Technology, Bayerischer Rundfunk, and the Australian Broadcasting Corporation

Aligning AI-powered personalisation with public-service values

News organisations are increasingly expected to personalise what audiences see. Stories can now be adapted into different formats, lengths, or versions depending on the user. At the same time, public service media organisations carry a responsibility to inform citizens about issues of public importance and to maintain a shared understanding of the world.

This project explores how public service media organisations can balance editorial commitments, audience preferences, and the possibilities of AI-driven news personalisation. It asks how personalisation can support, rather than undermine, editorial judgement, public-service values, and trustworthy news production.

From personalised content to shared public understanding

Personalised news raises urgent practical questions for people working in and around news production.

How should decisions be made when audience preferences, editorial judgement, and organisational values do not fully align? What happens when news content is adapted dynamically or with the help of generative AI? And how can editorial responsibility be maintained when different users may see different versions of the same story?

The project frames this challenge as one of pluralistic AI alignment: the need to design AI systems that can account for different, sometimes competing, values while preserving the public-service mission.

What we are doing

We are speaking with professionals across Europe and Asia-Pacific in editorial, product, technical, design, and strategy roles to understand how editorial values are interpreted, negotiated, and operationalised in everyday work.

The project also includes audience surveys, socio-technical modelling, and participatory design with newsroom professionals. The goal is to create practical tools, guidelines, and proof-of-concept systems for responsible personalisation in public service media.

Why this matters

Public service media organisations face a set of interconnected challenges that this project directly addresses.

  • News personalisation can make journalism more accessible and relevant, but it may also weaken shared public understanding.
  • AI-assisted content adaptation creates new uncertainty around editorial oversight, responsibility, and control.
  • Audience preferences, editorial judgement, and organisational values do not always align in everyday news production.

Underlying these challenges is a more fundamental question: how public service media can use personalisation responsibly while preserving editorial values, public trust, and its role in informing society.

Partners

Australian Broadcasting Corporation
Bayerischer Rundfunk
Queensland University of Technology

Team

Orestis Papakyriakopoulos

project lead

Cheng​ Yu​

LAB MEMBER

Fernanda Sauca

lab member