Futures with(out) Work organized a panel on “Algorithms, Truth and Justice” at the Gothenburg Science Festival. The discussion brought together Olgerta Tona, Charlotta Kronblad, and Aleksandre Asatiani to examine how algorithmic systems can contribute to injustice even when they are introduced with good intentions.
One core theme of the panel was that algorithmic decision-making is not only a technical matter. It also raises difficult questions about truth, accountability, and judgment. The panelists discussed why it can be hard to establish what counts as “true” in algorithmic decisions, particularly when systems rely on opaque models, biased data, or classifications in areas where the boundaries are not clear-cut to begin with.
To make these concerns concrete, the discussion turned to examples of algorithmic injustice. These included cases related to school placements in Gothenburg and the social welfare fraud detection system known as “robodebt” in Australia. Such examples helped illustrate how automated systems may amplify existing problems or produce harmful outcomes when complex social issues are translated into technical rules and indicators.
The audience engaged actively throughout the session. Questions focused on issues such as individual and organizational responsibility, as well as the motivations behind the growing use of algorithms in the public sector. These contributions added an important practical dimension to the discussion by highlighting that the use of algorithms is always embedded in institutional choices and governance arrangements.
The panel also addressed possible ways forward. Among the key points raised was the need to cultivate organizational cultures in which decision-makers are sensitive to the risks associated with algorithm use, especially in high-stakes contexts. The discussion also pointed to the importance of regulatory initiatives that can help create clearer safeguards around how algorithmic systems are designed, deployed, and evaluated.
Overall, the panel underscored that debates about algorithms are also debates about public values. Questions of justice, responsibility, and truth cannot be separated from the systems that increasingly shape important decisions.
The panel was supported by the RAISE project funded by Nordforsk.

