Reasserting the Human: Beyond Ethical Guardrails on AI Use and Development in Public Interest Journalism

Public interest journalism is an important pillar for protecting democracy. It is thus important to scrutinise the introduction of artificial intelligence (AI), a technology that comes into conflict with many underpinning values of the profession, across all stages of journalistic work. Current guardrails that aim to ensure ‘ethical use’ are codes of conduct or value-sensitive design, which aim to govern the use of AI by journalists and media organisations and the design process. However, neither guardrail captures the ontological reality of AI as a political assemblage of human and non-human elements. If we do not understand this assemblage as such, we cannot determine the impacts of the system on values and the overall position of journalists in society. In this theoretical contribution to the debate, we argue that instead of focussing on guardrails, collective efforts should aim at reasserting the agency of journalists who are best positioned to protect their profession. We offer three ideas for moving forward: First, to create more assemblage transparency by and for journalists, that is, transparency on all parts, actors, values, relations, and decisions that constitute the AI system. This assemblage transparency should involve the visualisation of power, which needs to go hand in hand with critical AI literacy, our second proposal. Critical AI literacy is the competence to ‘read’ how power and ideology manifest and are co-produced through AI. The final piece to reassert the agency of journalists must involve tackling the economic crisis of journalism.