Wikipedia@25: Knowledge at the crossroads

Just as human history is often divided into the eras B.C. and A.D.,it is no longer unreasonable to imagine the Internet’s story split between B.AI. and A.AI. — Before Artificial Intelligence and After Artificial Intelligence.In this dramatic transition, one quiet but revolutionary institution stands at a crossroads: Wikipedia, which turned 25 this January. Born in the age of dial-up modems and nurtured in the optimism of early cyberspace, Wikipedia now finds itself navigating the shadows — and possibilities — cast by AI.

When Wikipedia went online on January 15, 2001, it was the brainchild of two men who represented two contrasting intellectual temperaments of the digital age. Jimmy Wales, an internet entrepreneur with libertarian instincts, believed deeply in radical openness and crowd wisdom. Larry Sanger, a trained philosopher and Wikipedia’s first editor-in-chief, worried that unfiltered openness could undermine neutrality and epistemic rigor.Their collaboration lasted barely a year, but the tension between openness and authority, participation and expertise, democracy and discipline continues to define Wikipedia even at 25.Yet, against all skepticism, Wikipedia survived — and triumphed. At a time when knowledge was either locked behind academic paywalls or curated by elite institutions, Wikipedia demolished the idea that information must have gatekeepers.It declared, boldly and unapologetically, that knowledge has no monopoly. Anyone could write. Anyone could edit. Anyone could correct. In doing so, Wikipedia became the most radical experiment in the democratisation of information in human history.

Today, Wikipedia exists in more than 300 languages, hosts over 60 million articles, and is among the most visited websites in the world. For students in rural India, activists in Africa, journalists in Latin America, and citizens everywhere, Wikipedia became the first port of call on the information superhighway. It flattened hierarchies of knowledge dissemination and transformed passive consumers into active participants. The traditional flow of information — from elite producers to mass audiences — was disrupted forever.But Wikipedia’s silver jubilee arrives at a moment of profound uncertainty. Artificial Intelligence, particularly generative AI, threatens to redraw the architecture of knowledge itself. Tools powered by large language models no longer merely retrieve information; they synthesize, paraphrase, predict, and even “create” it. The question is no longer how humans access information, but whether humans will remain central to its production.Ironically, AI systems owe a significant debt to Wikipedia. Its freely licensed content has served as a foundational dataset for training AI models. In a sense, Wikipedia is both the parent and the prey of AI — nurturing the systems that may eventually overshadow it. Search engines already answer questions without directing users to Wikipedia pages. Chatbots summarize topics in seconds, eroding the incentive to consult original sources. The risk is not extinction, but invisibility.

This marks a shift from the age of information retrieval to the age of information mediation.Wikipedia was built for the former — a transparent, editable repository where users could see sources, debates, revisions, and disputes. AI, by contrast, often operates as a black box. It delivers confident answers without citations, context, or accountability. In doing so, it risks reversing one of Wikipedia’s greatest achievements: making the process of knowledge creation visible.At stake is not just Wikipedia’s relevance, but the very idea of democratic knowledge. Wikipedia’s model assumes that truth emerges through collective scrutiny, disagreement, and correction. AI models, however, optimize for coherence rather than truth, fluency rather than fairness. They reflect existing biases embedded in training data and amplify dominant narratives — often from the Global North, in English, and from digitally privileged communities.

This is where Wikipedia’s continued relevance becomes critical. In an AI-saturated ecosystem, Wikipedia remains one of the last large-scale, non-profit, human-curated knowledge platforms. It does not chase clicks, sell ads, or optimise for outrage. Its neutrality policy, though imperfect, is enforced through transparent rules and community deliberation. Errors are not hidden; they are corrected in public view. This ethos stands in sharp contrast to algorithmic opacity.

For countries like India, the stakes are even higher. Wikipedia has played a crucial role in vernacular knowledge dissemination, allowing histories, cultures, and perspectives outside Western academia to find space online. AI systems trained predominantly on English-language and Western-centric data threaten to marginalize these voices further. If Wikipedia weakens, so does the digital presence of the Global South.  AI can assist editors by detecting vandalism, translating articles across languages, summarising sources, and identifying gaps in coverage.

Wikipedia’s 25-year journey also offers a lesson for the broader politics of knowledge. In an era of misinformation, deepfakes, and epistemic chaos, trust has become the most valuable currency. Wikipedia earned trust not through authority, but through process — openness, verifiability, and collective responsibility.  AI systems, if they are to be trusted, must learn from this model rather than undermine it.

There is also a deeper philosophical question at play. Wikipedia represents the Enlightenment ideal that knowledge progresses through reasoned debate and shared inquiry. AI risks ushering in a post-Enlightenment moment, where answers are consumed without understanding and certainty replaces skepticism. In such a world, the danger is not ignorance, but uncritical acceptance.As Wikipedia steps into its second quarter-century, it must reassert its founding vision with renewed urgency.

The Internet may now be divided into B.AI. and A.AI., but Wikipedia must ensure that the human remains at the center of knowledge.

Its challenge is not technological obsolescence, but moral relevance.Twenty-five years ago, Wikipedia proved that knowledge does not belong to institutions, corporations, or experts alone. It belongs to humanity. In the shadow of Artificial Intelligence, that idea is more radical — and more necessary — than ever.