This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
Clarkesworld Magazine was forced to temporarily close its submission portal due to a massive surge in AI-generated spam stories. The influx, driven by 'get-rich-quick' schemes promoting ChatGPT, overwhelmed the magazine's editorial staff and threatened their open-submission policy. The editor noted that the volume of banned submissions increased from a baseline of 25 in October 2022 to 500 in February 2023, directly coinciding with the release of ChatGPT.
Sci-fi magazine Clarkesworld is reported to have temporarily stopped accepting submissions after receiving an overwhelming increase in purportedly LLM-generated submissions.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
Using AI systems to gain a personal advantage over others such as through cheating, fraud, scams, blackmail or targeted manipulation of beliefs or behavior. Examples include AI-facilitated plagiarism for research or education, impersonating a trusted or fake individual for illegitimate financial benefit, or creating humiliating or sexual imagery.
Also classified as
The flood of AI-generated content devalues human creative writing and disrupts the economic viability of short fiction markets that rely on open submissions.
Spammers used ChatGPT to cheat the submission system of Clarkesworld Magazine for personal financial gain, violating the publication's explicit rules against AI-generated content.
Human
Due to a decision or action made by humans
Intentional
Due to an expected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed
The incident was caused by human spammers intentionally using ChatGPT to generate and submit stories for financial gain after the model was deployed.
Added categories peak at 3: Substantial (Epistemic Harm, Inferred)
An influx of ChatGPT-generated submissions forced Clarkesworld Magazine to temporarily close its portal. While highlighting the challenges of managing mass-produced AI content, the incident is a localized commercial and moderation issue with negligible implications for national security.
The report describes ChatGPT as a tool that does not work autonomously, requiring a human to guide its output with a prompt that the AI model then attempts to automatically complete to assist in writing stories.
Limited Risk: ChatGPT is a chatbot and generative AI system. Under the EU AI Act, such systems pose a moderate risk and are subject to transparency obligations to ensure users are aware they are interacting with AI-generated content.
The AI system (ChatGPT) was used by spammers as a writing assistant to generate fictional stories for submission to Clarkesworld.
The reports are highly detailed and come directly from the affected editor (Neil Clarke) and reputable tech journalism (Ars Technica). They provide specific numbers, dates, and clear descriptions of the operational impact on the magazine.
The incident could be viewed purely as a spam/moderation challenge rather than an AI safety issue, but the central role of generative AI capabilities makes it a clear AI safety and dual-use risk.
Technology
Generative AI Availability
ChatGPT allows rapid generation of text, enabling mass spam creation.
Unreliable Detection Tools
AI text detectors are inaccurate, causing false positives and negatives.
Data Inputs
Dishonest Submission Inputs
Submitters lie about using AI on forms, bypassing guidelines.
Plagiarized and AI Text
Submissions contain machine-generated or plagiarized stories.
Human Factors
Get-Rich-Quick Motivation
Spammers are driven by online claims of easy money using ChatGPT.
Disregard for Rules
Submitters ignore the magazine's explicit ban on AI-assisted work.
Process and Methods
Always-Open Submissions
Continuous open submission windows attract higher volumes of spam.
Manual Review Bottleneck
The process relies on manual reading, which cannot scale with AI volume.
Management
Inability to Fund Tools
The magazine cannot afford expensive third-party identity verification.
Commitment to Open Access
Refusal to use barriers like pay-to-submit that exclude new authors.
AI system involved
Identified by automated extraction; can differ from the alleged developers above.
