In recent years, institutions across sectors — from publishing to government and academia — are being inundated by text created by generative artificial intelligence, creating an escalating challenge for traditional systems that once relied on human effort to regulate and limit volume.
The trend first became widely notable in 2023, when the science fiction magazine Clarkesworld temporarily stopped accepting submissions because an overwhelming number were produced with AI tools rather than by human authors. Many contributors reportedly fed the magazine’s submission guidelines into AI and submitted the output as original work. Other literary outlets have reported similar spikes in AI-generated entries.
This phenomenon isn’t isolated. Newspapers are seeing surges of AI-generated letters to the editor; academic journals are receiving large numbers of machine-written papers; lawmakers are flooded with AI-crafted constituent comments; and legal systems globally are coping with a torrent of filings from self-represented litigants using AI to draft legalese. Social media platforms, scientific conferences and even hiring and education systems are contending with the same pattern of automated content volumes.
Many institutions initially responded by tightening submission processes or suspending them entirely. Others have turned to defensive uses of AI. For example:
-
Academic peer reviewers increasingly rely on AI tools to help flag suspect papers;
-
Social platforms deploy AI moderators to lessen harmful or bogus posts;
-
Courts use AI triage systems to manage burgeoning case filings;
-
Employers and educators use AI to screen applications and support grading or feedback.
However, this has created what experts describe as an “arms race”: as AI is used to generate content, organizations use other AI tools to detect and filter it. These systems evolve rapidly but are inherently adversarial — for every detection improvement, generative AI adapts, making perfect identification essentially unattainable.
The effects of this arms race are mixed. On one hand, society risks significant disruption if courts are clogged with frivolous AI-manufactured cases or if academic merit is distorted by mass-produced machine writing. On the other hand, AI assistance also democratizes access to writing support that historically favored the affluent — for instance, helping researchers for whom English is a second language or enabling broader participation in discourse.
Yet the fundamental difficulty remains: there is no reliable way to fully distinguish human from machine text, and institutions cannot simply “turn off” the technology. The explosion in AI output means organizations must adapt their workflows, embrace assistive technologies, and balance the benefits of AI-augmented content with safeguards against fraud, misinformation and overload.
The ongoing escalation suggests that rather than containing AI text, society will have to learn to coexist with it — rethinking how we validate, curate and trust written content in an age where machine-generated text is ubiquitous.
The authors do not work for, consult, own shares in or receive funding from any company or organization that would benefit from this article, and have disclosed no relevant affiliations beyond their academic appointment.
This article was originally published on The Conversation.
