The study examined English-language newspaper sources from the United States and Europe up to December 31, 2024, assembling a final corpus of 43,027 articles, of which 9,684 came from the United States and 33,343 from Europe. The authors combined computational text analysis with interpretive examination of the corpus, a methodological pairing that allowed them to detect broad patterns of representation across a very large body of coverage while also reading individual articles closely for nuance. The scale of the dataset matters: with tens of thousands of articles spanning years of reporting, recurring patterns of framing and sourcing become statistically visible in ways that smaller samples cannot reveal.

What emerged from the analysis is what the researchers describe as a striking hierarchy of visibility and authority. Massimo Ragnedda, professor of media and communication at the University of Sharjah and the study’s senior author, said the findings reveal that the United States is predominantly positioned as a technological innovator and leader, while Europe is cast as a regulator and normative evaluator. In this symbolic geography, America supplies the vision of the AI future and Europe supplies the moral and legal guardrails, a division of narrative labor that mirrors but also reinforces real-world concentrations of technological and institutional power.

China’s position in the coverage is more complicated and more ambivalent. Although Beijing is a leading AI power alongside the United States, the authors found that it appears as a hyper-visible but ambivalent rival, associated in the reporting with surveillance, geopolitical competition, and strategic threat. In other words, China is never ignored—its AI ambitions receive enormous attention—but that attention is filtered through frames of suspicion and rivalry rather than through the celebratory innovation narratives routinely applied to American firms and researchers. Visibility, the study suggests, is not the same as favorable representation.

Most striking of all is who is missing. Countries in Africa and Latin America are largely absent from the narrative or positioned primarily as sites of implementation, risk, or developmental need. They appear, when they appear at all, not as authors of technological futures but as territories where other people’s technologies are deployed, where risks materialize, or where development gaps must be closed. The result is a global AI story told almost entirely from a narrow band of the English-speaking world, in which the majority of humanity figures as a backdrop rather than as a protagonist.

The researchers conceptualize this pattern as discursive stratification, a term meant to capture an unequal distribution of narrative authority in which some actors are repeatedly positioned as legitimate authors and interpreters of technological futures, while others appear mainly as recipients, sites of implementation, or objects of technological change. The metaphor of stratification is deliberate: like geological layers, the media coverage sorts nations into tiers of narrative standing, with the top layers granted the authority to imagine and define what artificial intelligence will become, and the lower layers assigned passive or problematic roles in that unfolding story.

This framing draws on recent political-economy scholarship that views AI not simply as an unprecedented technological innovation with far-reaching consequences for human life, but as a sphere beset by rivalries over power distribution, geopolitical competition, and infrastructural concentration. The study extends that perspective into the realm of communication, treating AI as an inequality regime in which communicative power, visibility, value, and opportunity are redistributed through unequal linguistic structures of control. On this account, the struggle over AI is fought not only in laboratories, data centers, and regulatory agencies, but in the language used to describe the technology in the world’s most influential newspapers.

Central to the argument is a redefinition of what news does. The researchers conceptualize news not as a mirror of technological change, but as a narrative infrastructure—a discursive and institutional system through which AI becomes publicly legible and through which some actors are positioned as innovators, regulators, threats, or absences. Just as physical infrastructure such as undersea cables and semiconductor supply chains determines who can access and deploy the technology, this narrative infrastructure determines who can meaningfully participate in the public conversation about it. The stories repeatedly told about AI, the authors argue, help shape perceptions of who leads, who regulates, who threatens, who benefits, and who remains largely unheard.

The implications extend well beyond media criticism. Ragnedda notes that AI does not simply determine what we talk about; it also helps determine who is recognized as having the authority to speak about the technology and whose futures are considered worth imagining. If the public understanding of artificial intelligence is filtered through a narrow set of voices, then policy debates, investment decisions, and regulatory frameworks built on that understanding risk inheriting the same blind spots. The absence of African and Latin American perspectives, for example, means that questions of digital colonialism, labor exploitation in data annotation, and unequal access to AI benefits may be systematically underexplored in the outlets that most shape elite opinion.

The authors are nonetheless optimistic about the study’s contribution to the field. They write that it advances communication scholarship beyond descriptive analyses of representation toward a structural account of narrative authority in global media systems, showing how framing, sourcing hierarchies, and thematic emphasis combine to produce the symbolic geography of AI. Ragnedda distills the stakes into a single question: the issue is not only who builds AI, but who gets to define what AI means, and whose voice is missing. As artificial intelligence continues to reshape economies, security doctrines, and daily life, the study suggests that paying attention to the chain of dominance and dependence embedded in everyday news coverage may be as important as worrying about the technology itself.

Subject of Research: Discursive stratification and narrative authority in Western English-language news coverage of artificial intelligence

Article Title: Though seen as a threat to humanity, AI reporting in the West reflects a clear chain of dominance and dependence

Article References: Though seen as a threat to humanity, AI reporting in the West reflects a clear chain of dominance and dependence. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: artificial intelligence, media narratives, discursive stratification, narrative authority, news media, United States, Europe, China, Global South, computational text analysis, communication research, geopolitical competition

The research team, led by Nick Pietsch, Naomi Baes, and Nick Haslam of the University of Melbourne, applied a framework known as SIBling, which measures semantic change along three dimensions: Sentiment, Intensity, and Breadth. Sentiment captures the average valence of words appearing near the target term, Intensity captures their average emotional arousal, and Breadth captures the diversity of semantic contexts in which the term is used. The approach builds on a body of work showing that harm-related concepts such as trauma, addiction, and mental health have broadened their meanings over recent decades, encompassing less severe phenomena and more varied situations. Whether a diagnostic label like ADHD follows the same trajectory had not previously been tested at scale.

The corpus was drawn from the U.S. Newsstream Collection, a ProQuest database of more than one billion documents from over 1,300 sources. Articles were selected if they contained ADHD, A.D.H.D., attention deficit hyperactivity disorder, or attention deficit disorder in their body text, then de-duplicated using a shingling and Jaccard similarity procedure with a threshold of 0.6. The final corpus contained 63,963 articles, from which 103,532 clean sentences containing the target term were extracted, spanning 191 publishers. Coverage grew dramatically over the study period, from just 47 articles from 11 publishers in 1990 to 2,334 articles in 2023, with newspapers and wire feeds making up more than 90 percent of the material.

To measure Breadth, the researchers sampled 50 ADHD-containing sentences per year, embedded each one using XL-LEXEME, a target-aware sentence embedding model fine-tuned to distinguish whether a word expresses the same or different senses across sentence pairs, and computed the average pairwise cosine distance between embeddings. Higher distances indicate that the term appears in more heterogeneous contexts. The result was a statistically reliable upward trend, equivalent to roughly an 8 percent increase in Breadth across the 36-year window. Crucially, when the team modeled whether rising salience, indexed by the growing proportion of articles mentioning ADHD, predicted subsequent semantic change using autoregressive distributed lag models, only Breadth showed a significant lagged effect of salience. In other words, the more the media talked about ADHD, the more loosely and variously the term was used in the years that followed.

The Sentiment and Intensity results were more surprising. The researchers expected that growing familiarity with ADHD, the normalization of diagnosis and treatment, and the rise of neurodiversity discourse would make coverage more positive and less emotionally charged. Instead, the average valence of words within five lemmas of ADHD declined by about 0.34 points on a nine-point scale, a modest but highly reliable drop, while average arousal rose by about 0.07 points. Both trends survived extensive sensitivity checks: they held when the context window was widened to the full sentence, when an alternative affective lexicon, the NRC Valence, Arousal, and Dominance Lexicon, was substituted for the Warriner norms, and when the partial year 2025 was excluded from the analysis.

A corpus-level benchmark helped clarify what these trends mean. The researchers randomly sampled 1,000 articles per year from the broader database and computed their average valence and arousal. General news language showed no long-term decline in valence, meaning that the negative drift around ADHD was specific to the condition rather than a secular shift toward gloomier journalism. The rise in arousal, by contrast, did not significantly exceed the corpus-wide increase, suggesting that ADHD coverage became more emotionally intense largely in step with news language generally, though a late reversal in the ADHD arousal trajectory after 2023 was distinctive.

What drove the drift toward darker associations? A contribution analysis of the words neighboring ADHD pointed to two complementary changes. Positively valenced childhood terms, especially child and parent, became steadily less frequent, while negatively valenced clinical and mental health terms, including autism, anxiety, depression, medication, and symptom, became more common. The ten largest trend-consistent contributions accounted for roughly 37 percent of the valence decline. The pattern suggests that news coverage has shifted from framing ADHD as a childhood behavioral concern toward framing it as one node in a wider web of psychiatric conditions, a shift that both broadens the concept and subtly problematizes it. The rise in arousal was more diffuse, reflecting both the disappearance of low-arousal filler words and the arrival of higher-arousal terms such as depression, bipolar, risk, and disorder.

The study also detected a striking nonlinear turn in the late 2010s. Sentiment, which had declined since 1990, began rising after its low point in 2018, and Intensity began falling after its peak in 2019. The researchers offer three speculative explanations that coincide with this inflection: the COVID-19 pandemic, which brought a surge of media attention to ADHD that may have been more sympathetic than earlier coverage; the international explosion of TikTok in 2018, a platform now notorious for both ADHD awareness content and misinformation; and the mainstreaming of neurodiversity discourse, which reframes ADHD in terms of strengths like hyperfocus and creativity. A parallel study of British newspaper coverage of autism found a similar shift from deficit-focused to diversity-focused language around 2018, lending plausibility to the third explanation, though the authors caution that disentangling these influences awaits further work.

The implications extend beyond lexicography. Because news media are a major vehicle of cultural transmission, shifts in how ADHD is represented plausibly shape public attitudes and behavior. Research by Tse and Haslam has shown that people who hold broader concepts of mental disorder self-diagnose at higher rates, at lower levels of distress, and seek treatment more often, while other work suggests broad definitions lead people to judge others as less likely to recover. Rising awareness itself has been theorized to fuel a prevalence inflation, in which mild distress is overinterpreted as illness. Against that backdrop, the finding that ADHD has expanded its semantic range precisely as its media salience soared, and that coverage has grown somewhat more critical in tone, offers a quantitative thread connecting diagnostic expansion, cultural attention, and the well-documented rise in self-diagnosis and treatment-seeking.

The authors are careful about limits. The trends, though statistically robust, are modest in absolute magnitude, and the lexicon-based measures capture local lexical associations rather than the overall tone of complete articles. The corpus is restricted to US media, early-year coverage was thinner, and some articles using the unpunctuated abbreviation ADD may have been missed. Still, the scale of the analysis, more than 60,000 articles over 36 years, dwarfs previous media studies of psychiatric language, and its demonstration that a diagnostic concept can creep in public discourse opens a new window on how mental health categories live and grow outside the clinic. Whether the recent softening of sentiment and intensity marks a durable turn toward normalization, or merely a pause in a longer drift toward problematization, is now one of the most consequential questions the trend lines pose.

Subject of Research: Historical semantic change in US news media representations of ADHD

Article Title: Shifting focus: Changing representations of ADHD in US news media

Article References: Pietsch, N., Baes, N., & Haslam, N. (2026). Shifting focus: Changing representations of ADHD in US news media. SSM – Mental Health, 10, Article 100710. https://doi.org/10.1016/j.ssmmh.2026.100710

Image Credits: AI Generated

DOI: 10.1016/j.ssmmh.2026.100710

Keywords: ADHD, concept creep, semantic change, news media, natural language processing, word embeddings, sentiment analysis, mental health, diagnostic inflation, neurodiversity, corpus linguistics, SSM Mental Health