Journalists are supposed to hold power to account, and they’re deeply uneasy about who controls AGI.
That’s the key finding from the inaugural Spark Hunter Survey, which canvassed a group whose views on AGI have never been systematically explored before: the reporters covering the field.
The survey sought insights from 31 science and technology journalists with diverse backgrounds. Over more than 50 questions, from open-ended prompts to Likert-type items, their responses covered AGI’s development, risks, governance, and the role of the media in shaping public understanding.
Overall, the respondents see AGI as potentially transformative, but being shaped by institutions they don’t fully trust. They worry that it’s driven by commercial incentives, powerful actors, and opaque processes — and that the general public isn’t prepared for the impacts.
A trust problem
At the core of their concerns is the issue of motive. An overwhelming 87.1% of journalists view economic profit and market dominance as the primary motivations behind the current AGI push.
Scientific curiosity and the pursuit of knowledge came far behind, at 35.5%, while only 22.6% selected advancement of human wellbeing. To the reporters covering the beat, commercial interests dramatically outweigh more idealistic motivations.
That skepticism about motive is compounded by a lack of faith in the industry’s honesty. Most journalists have either no trust at all (41.9%) or very little trust (35.5%) in AGI developers being transparent about their progress and capabilities.
This distrust extends to their own profession. Only 3% think that current media discussions about AGI are balanced and objective, while the same paltry percentage thinks there’s sufficient critical and investigative journalism regarding AGI claims and development.
“All too often, publications echo wild speculation from major AI firms about the technology’s supposed godly qualities,” says Tamlin Magee, a freelance journalist based in the UK, in commentary sent to AGI Ethics News.
Power and control
When conversation turns to who actually controls AI, the distrust becomes alarming. More than half of respondents (54.8%) said they were “extremely concerned” about AGI being developed by a small number of powerful entities (corporations or nations).
That concern is closely linked to doubts about humanity’s ability to manage increasingly capable systems. When asked how confident they were that humanity could control AGI if it significantly surpassed human intelligence, 45.2% said they were “not at all confident.” Only 22.6% selected either “very confident” or “extremely confident.”
Together, these findings show that journalists believe power over AGI is consolidating in too few hands, and that those hands may not be able to hold on once the technology is built.
“If LLMs are guilty of being Stochastic Parrots, some reporters are surely Eschatological Parrots,” Magee says. “The evolution of AI is not some predestined path to artificial general intelligence or the recalibration of the world under the enlightened Silicon Valley few. By repeating doomsday scenarios, they lend a mystical, obfuscating air to what are ultimately technologies under human control.”
To address this, society requires greater scrutiny of AI’s powerbrokers — a task well-suited to reporters. After all, a machine’s impact depends not only on its capabilities, but also what humans choose to permit or restrict.
The threat landscape
The survey also reveals where journalists believe the most immediate dangers lie.
Asked for their three biggest concerns, most prioritized practical risks. Their most common concern is misuse by malicious people, which 80.6% selected, followed by severe economic inequality (51.6%) and erosion of human skills and capabilities (38.7%). Trailing in the distance were the speculative threats of an existential risk to humanity (19.4%) and unforeseen “black swan” catastrophic events (12.9%), despite both receiving extensive media coverage.
This suggests that technology journalists aren’t fixated on futuristic science-fiction scenarios. Yet this practical skepticism doesn’t always survive contact with the page. As Magee notes, exaggerated narratives around AGI can obscure more pressing concerns.
“If LLMs are guilty of being Stochastic Parrots, some reporters are surely Eschatological Parrots,” he says. “The evolution of AI is not some predestined path to artificial general intelligence or the recalibration of the world under the enlightened Silicon Valley few. By repeating doomsday scenarios, they lend a mystical, obfuscating air to what are ultimately technologies under human control.”
Uses in war are another cause of alarm. A combined 83.9% of respondents agreed or strongly agreed that the development of AGI for military applications is inevitable, accentuating the importance of who controls the technology.
The governance gap
None of these concerns curdle into fatalism. Instead, the journalists collectively expressed strong support for proactive governance.
More than three-quarters of respondents (77.4%) rated international treaties and agreements on AGI research and deployment as very or critically important. Global monitoring systems for AGI capabilities and proliferation received similar backing, at 74.2%. Interdisciplinary collaboration between technologists, ethicists and social scientists was also viewed as near essential, with 70.9% rating it very or critically important.
These findings underscore the need for a new reporting agenda. Treaties, monitoring regimes, and governance proposals rarely get the column inches that a splashy model launch does, even though many journalists rate them as more urgent. The journalists surveyed don’t just want more coverage of AI capabilities; they want more scrutiny of the institutions, rules and incentives shaping those capabilities.
But perhaps the biggest warning for journalists is that society is struggling to keep up. Asked whether they feel that the current AGI discourse is sufficiently preparing the public for its potential arrival, 77.4% of respondents said no.
That places journalism at the center of the challenge. If the people covering AGI believe the public is underprepared, they need to provide better coverage. Only then can we hold the people who will control AGI accountable.
Ethical concerns addressed in this article:
- Who should have access to AGI technologies, and how do we prevent their monopolization by wealthy corporations or governments?
- What ethical considerations arise from the use of AGI for military or security purposes?
- What mechanisms can be developed for effective, democratic oversight of AGI, and how is the public meaningfully engaged in decisions about its deployment and governance?
Thomas is a freelance tech reporter who has covered AI for nearly a decade. He was previously managing editor of The Next Web, a Financial Times company. In his free time, he likes playing chess (badly) and basketball (even worse). You can reach him at thomas.s.r.macaulay@gmail.com