Wikipedia’s 8% Pageview Slide: WMF blames AI search answers & social video; Google says clicks are fine

# Wikipedia’s 8% Pageview Slide, and what it means for AI in Morocco Wikipedia just reported a notable shift in how people find and consume knowledge. The Wikimedia Foundation (WMF) says human pageviews fell about 8% year-over-year, after reclassifying a mid-2025 traffic surge as bots. WMF points to generative-AI answers in search and social video habits as the key drivers. Google counters that overall click volume is relatively stable and more links appear in AI Overviews. This debate matters in Morocco. The country is building an AI ecosystem across startups, universities, and public services. Discovery patterns will influence funding, participation, and the visibility of local knowledge. ## What changed in the data, and why the 8% matters Around May 2025, WMF saw a spike in seemingly human visits, mostly from Brazil. Bot-detection systems were updated, and traffic for March–August 2025 was reclassified. Much of the spike turned out to be bots designed to avoid detection. Without the inflated visits, WMF sees a real drop in human pageviews versus the same months in 2024, roughly 8%. WMF links the decline to AI-generated answers that satisfy queries without a click. Younger users are also moving toward social video for quick facts. Fewer site visits can reduce the flow of new contributors and small-donor revenue. That risks the sustainability of open, verifiable knowledge. ## Morocco’s discovery shift is already visible Moroccan users increasingly get summaries in search and short videos on platforms. Many queries resolve without a visit to a source page. That changes how publishers, educators, and startups reach audiences. Students and professionals still rely on Wikipedia for depth. Yet habits favor instant answers and short formats. If links are bypassed, Moroccan knowledge creators lose feedback loops and community growth. ## Morocco’s AI ecosystem, in brief Morocco’s startup scene is growing around applied AI. Founders focus on practical problems in agriculture, health, logistics, and customer service. Incubators and innovation hubs help turn prototypes into businesses. Technopark sites in Casablanca, Rabat, and Tangier host digital startups. UM6P’s StartGate supports founders with labs, mentors, and access to industry partners. Public and private institutions are opening data and modernizing infrastructure. ## Startups delivering AI use cases Atlan Space builds autonomous drone software that helps monitor oceans and protected areas. Its systems can patrol large zones and flag anomalies for authorities. The company has worked with African partners on environmental and fisheries missions. It represents Morocco’s edge in applied autonomy. Sowit supports precision agriculture with analytics that aid irrigation, fertilization, and yield planning. Farmers use insights to cut resource waste and improve returns. These tools align well with Morocco’s water and sustainability priorities. They show AI’s practical value in the field. DataPathology digitizes pathology workflows and supports remote diagnostics. AI assistance can help triage slides and streamline case management. Hospitals and labs gain speed and resilience. The model suits Morocco’s distributed healthcare landscape. ## Universities, infrastructure, and talent UM6P runs advanced computing infrastructure and hosts data science programs. Researchers and startups collaborate on applied machine learning and modeling. Universities in Rabat, Casablanca, and other cities expand AI curricula. Local teams work on Arabic, Darija, and Amazigh language technologies. Projects target text classification, translation, and OCR. That helps build tools rooted in Morocco’s linguistic reality. ## Government initiatives and open data foundations Morocco’s Agence de Développement du Digital (ADD) drives digital transformation and ecosystem support. Public agencies are building services that rely on data and automation. The government maintains open data portals that feed civic and business innovation. Large events bring partners together. GITEX Africa has hosted AI tracks and startup showcases in Marrakech. These gatherings help align policy, industry, and research. ## Practical AI uses in Morocco today Public services use analytics for planning and logistics. Trade and customs processes benefit from automation and risk scoring. Municipal operations test predictive insights for maintenance and service delivery. Banks deploy machine learning for fraud detection and customer support. Telecom operators optimize networks with AI-driven analytics. Tourism services explore chatbots that guide visitors using trusted sources. Small businesses add AI to support sales and operations. Common tools include content assistants, FAQ chatbots, and basic recommendation systems. These efforts improve the customer experience with modest budgets. ## WMF’s asks, and Morocco’s role WMF welcomes new access methods but stresses attribution and traceability. The content powering answers comes from human editors. WMF wants platforms to drive more visitors back to Wikipedia and make attribution obvious. Users should easily trace claims to sources. To support this, WMF is building an attribution framework and enforcing responsible reuse via Wikimedia Enterprise. It is investing in two reader-focused teams: Reader Growth and Reader Experience. A Future Audiences program is experimenting with Wikipedia on YouTube, TikTok, Roblox, and Instagram. WMF is also testing video, games, and chatbots to meet new habits. Moroccan publishers and creators can align with these goals. Build experiences that show citations clearly and invite clickthrough. Bring human editorial context forward, especially for sensitive topics. ## Google’s counterpoint, and a pragmatic read Google says organic click volume is relatively stable year-over-year. It argues click quality is higher and AI Overviews surface more links. The company claims traffic is redistributing toward forums, video, and first-person content. It emphasizes that sites retain control via open-web protocols. Both views can be true in Morocco. Some verticals may gain clicks while reference sites lose a share. AI answers and social video compress discovery for casual queries. Deeper content still wins when context and credibility matter. ## Guardrails for AI features on content sites WMF paused a pilot of AI-generated summaries earlier in 2025. Editors raised accuracy and credibility concerns. The lesson is clear: AI layers must strengthen verifiability, not dilute it. Moroccan sites should apply similar guardrails. Use models to assist humans, not replace them. Make sources, timestamps, and update logs visible. ## What WMF is doing next, and how Moroccans can help WMF is recruiting volunteers to test reader experiments. It is improving mobile editing and growing the next generation of advanced editors. It also asks users to click through to sources and support human-curated knowledge. Everyone can help shape new reading experiences. Moroccan communities can participate in these efforts. Engage with local Wikimedia chapters and meetups. Contribute to editing workshops in universities and public libraries. Bring Darija and Amazigh content into verifiable, accessible formats. ### Action ideas for Morocco - Publishers: Add clear citations, provenance labels, and schema metadata. Invite clickthrough with explainer boxes and “read at the source” nudges. - Publishers: Track referral changes from search, social, and AI assistants. Adjust content packaging for short, deep, and visual formats. - Educators: Teach source evaluation and citation tracing in curricula. Show students how to move from summaries to primary references. - Educators: Use Wikipedia as a training ground for research and editing. Reward contributions that improve local pages and citations. - Startups: Build products that respect licensing and attribution. Provide users with one-click paths to underlying references. - Startups: Offer bilingual UX in Arabic, French, Darija, and Amazigh. Leverage local language models where available. - Platforms and creators: Add visible credits in videos and chatbots. Link directly to relevant Wikipedia pages and primary sources. - Platforms and creators: Pilot with WMF’s Future Audiences ideas. Test short-form explainer series that invite deeper reading. ## Case flows that serve users and the open web A tourism chatbot can provide concise guidance and link to detailed pages. It can cite Wikipedia and official ministry sources for history and rules. Users get fast answers and clear pathways to trusted context. An agriculture assistant can summarize irrigation tips and cite research. It can link to local extension services and open datasets. Farmers gain practical advice and traceable references. A classroom bot can field questions in Darija and point to cited articles. It can encourage students to check edit histories and references. That builds habits of verification and participation. ## Key takeaways - AI answers and social video are reshaping discovery, including in Morocco. - Wikipedia’s 8% human pageview drop risks editor and donor pipelines. - Attribution and traceability are crucial to sustain open knowledge. - Morocco’s AI ecosystem can thrive by respecting sources and inviting clicks. ## Closing note TechCrunch covered WMF’s analysis and Google’s response on October 18, 2025. The AI answers versus open-web clicks debate is now mainstream.