Wikipedia bans ai-generated articles after editors vote to protect human knowledge
Wikipedia's army of volunteer editors has drawn a hard line in the sand: no more ai-generated content on the world's largest encyclopedia. The ban, approved by an overwhelming majority after months of heated debate, marks the most aggressive stance yet by a major platform against the rising tide of algorithmic text.
The 20-year knowledge project fights back
Since 2001, Wikipedia has relied on human expertise to build its 60-million-page repository of knowledge. But the emergence of large language models has flooded the platform with machine-written entries that often read convincingly while containing subtle errors or fabricated sources. Editors call it 'ai slop' – the digital equivalent of junk food, palatable but nutritionally worthless for knowledge.
The new guidelines, slipped into Wikipedia's editing policies last week, prohibit using generative ai for creating, expanding or editing articles. The move follows internal data showing a 200% spike in ai-generated submissions since ChatGPT's public launch, with many entries containing hallucinated citations or plagiarized content scraped from copyrighted sources.
What ai can and cannot do on wikipedia
The ban isn't absolute. Editors can still deploy ai as a sophisticated spell-checker, using it to catch grammatical errors or suggest stylistic improvements – provided the algorithm adds zero original content. Translation assistance remains permitted, but only for editors who can verify the AI's output against source materials in the original language.
The distinction reveals Wikipedia's calculated gamble: embrace AI's utility while rejecting its authorship. It's a surgical approach that platforms like LinkedIn and Medium have struggled to implement, often settling for disclosure requirements that few users read.
Behind the servers that nearly broke
The policy shift arrives as Wikipedia's infrastructure faces unprecedented strain. AI companies have been scraping the entire encyclopedia multiple times daily, with bots sending up to 10,000 requests per second to download complete database dumps. The platform has brokered deals with major AI firms to provide pre-formatted data feeds, protecting its servers from collapse while maintaining free access for legitimate researchers.
Traffic data obtained by TechBloom shows scraping attempts increased 1,500% since 2022, with some days seeing Wikipedia hit harder than during its peak viral moments. The encyclopedia's operating budget, a lean $165 million annually, simply cannot absorb the computational cost of serving both human readers and AI training pipelines.
The editor consensus is clear: Wikipedia's value lies not in accumulating information, but in the human judgment that curates it. Twenty years of meticulous cross-referencing and fact-checking created a resource that AI models now mine for training data. By banning generative AI from contributing, Wikipedia protects both its servers and its soul.