A recent study published in the journal Organization Science reveals that artificial intelligence is flooding academic publishing with lower-quality work. While AI tools offer benefits such as summarizing research and improving writing mechanics, the study warns they have triggered a surge of poorly written submissions and reviews. The authors state that AI language models, combined with intense publish-or-perish incentives, are driving the academic field toward quantity over quality. The journal, a leading publication in social sciences, received papers from major universities and research teams globally. Concerned about the erosion of standards, an AI task force composed of its editors conducted a comprehensive review. The team analyzed nearly 7,000 submissions and over 10,000 reviews spanning from 2021 to 2026. Starting the study in 2021, two years before the public launch of ChatGPT, allowed researchers to accurately compare writing quality before and after the widespread adoption of AI. To detect AI usage, the task force employed the Pangram AI detection tool, which identifies characteristic writing traces. Each document was scored from 0 for entirely human-written content to 1 for fully AI-generated text. In addition to analyzing published papers, the team examined all submitted drafts and private peer reviews. They also utilized standard tests to measure readability and style, such as the Flesch Reading Ease score. The results indicate a dramatic shift following the arrival of ChatGPT. The volume of submissions increased by 42%, a trend largely attributed to AI assistance. By early 2026, the majority of manuscripts incorporated AI to some degree. However, this rise in volume correlated with a decline in writing quality, making papers increasingly difficult to read. The study identified two specific groups most likely to use AI: research teams from non-native English-speaking institutions and new entrants to the field with limited submission experience. Interestingly, using AI was associated with higher rejection rates. Academics at top business schools and those under significant pressure to publish also showed a marked increase in AI-assisted work. The impact of AI extends beyond authors. More than 30% of expert reviews submitted to the journal now involve language models, a sharp rise from previous years. The task force noted that these AI-generated reviews tend to be narrower and less insightful than those produced by humans. This trend places additional strain on editors, who must now invest significant time filtering out low-quality submissions and reviews. In response to these challenges, the journal suggests a fundamental overhaul of how academic research is valued. The study argues that the focus should shift from the sheer number of papers published to the quality of the ideas presented. This shift aims to mitigate the pressure that currently drives researchers to rely on AI for speed rather than substance, thereby preserving the integrity of the peer-review system.