In: Information, Communication & Society |DOI|

How do the media in the Global North and the Global South frame artificial intelligence (AI)?

No previous technology reached 1.2 billion users worldwide in as little time as artificial intelligence. This diffusion, however, is profoundly unequal, with an adoption rate in the global North roughly twice as high as that recorded in the South. Previous research has analysed the economic dimensions of this asymmetry, such as infrastructure, labour, resource extraction and environmental costs. Little was known, however, about how these inequalities are reflected in media representations concerning AI priorities, risks and governance agendas.

It is this gap that Bárbara Mota's article, published in the journal Information, Communication & Society, helps to fill. After analysing 6,829 articles published between 2010 and 2024 in twelve major newspapers, six per region, combining topic modelling and sentiment analysis, the study shows that North and South occupy distinct stages in the process of technological appropriation and, for this reason, produce fundamentally divergent narratives about AI. The research also identifies two decisive moments in the politicisation of the technology. The first occurred in 2018, following the Cambridge Analytica scandal, which highlighted the potential of algorithms to influence electoral processes. The second came with the launch of ChatGPT in 2022, when journalistic coverage of AI and politics more than sextupled in just two years, driven by debates on large scale disinformation, electoral impacts and regulation.

Another finding further evidences the growing weight of geopolitics in AI coverage. The technological rivalry between China and the United States appears as one of the main narrative axes, frequently described in terms of a technological race, strategic competition and national security.

The findings engage with recent debates on global AI governance. Last month, economist Daron Acemoglu argued that developing countries stand to lose the most from an AI agenda centred on automation. Mota's article helps to show that this same world is also the one that places the greatest hopes in the technology, especially in its potential to reduce historical asymmetries of technological dependence and accelerate development trajectories.

For the author, rather than simply reporting the news, media representations constitute one of the terrains of dispute where the future of AI is being paved. Regulating AI and defending technological sovereignty therefore begins with understanding what perceptions different societies are constructing about it.