Algorithms Now Run Science: New Study Reveals AI Controls Funding & Publishing
A new study shows that artificial intelligence algorithms are secretly controlling which scientific experiments get funded and which findings get published, threatening to stifle breakthrough ideas.
A study published in the July 2026 issue of *Science* (Volume 393, Issue 6806) reveals that artificial intelligence algorithms are quietly shaping the direction of global scientific research [188114]. The report warns that these digital “invisible hands” now decide which experiments get funded and which findings are published, creating a feedback loop that favors studies matching past successful patterns over novel ideas [188114].
Researchers analyzed thousands of grant applications and journal submissions, finding that AI systems originally designed to speed up data analysis have begun influencing peer review and funding decisions [188114]. The authors warn that without oversight, science risks becoming a closed system where novelty is filtered out before it ever reaches a human reviewer [188114]. They call for transparency and urge scientific bodies to audit these algorithms regularly [188114].
In the same issue, another paper proposes a “classroom contract” for AI, outlining a framework to govern how AI systems interact with students and teachers [188093]. The contract aims to establish clear boundaries for AI use in education, ensuring transparency and accountability while preventing misuse and encouraging innovation in learning environments [188093].
A separate initiative published in the same issue aims to institutionalize AI diplomacy across Africa, creating formal government bodies and policies to manage international negotiations and cooperation on AI [188110]. The plan seeks to help African nations address challenges like data sovereignty, algorithmic bias, and the digital divide, giving the continent a stronger, more unified voice in global tech governance [188110].