The Wikimedia Foundation, which operates the online encyclopedia, said on Thursday that it has also signed agreements over the past year with AI-focused companies such as Perplexity and France’s Mistral AI. Meta and Amazon were already partners, while Google’s parent company Alphabet, entered into a similar arrangement with Wikimedia in 2022.

The move comes after Wikimedia repeatedly raised concerns about technology companies using Wikipedia’s freely available content to train AI models without contributing financially, a practice it said has increased server traffic and operating costs. Wikipedia hosts more than 65 million articles across over 300 languages, making it one of the most important sources of training data for generative AI tools developed by major tech firms.

Wikimedia has been promoting wider adoption of its Wikimedia Enterprise product, which allows companies to pay for structured, large-scale access to its content tailored for AI training and commercial use.

“Wikipedia is a critical component of these tech companies’ work that they need to figure out how to support financially,” said Lane Becker, president of Wikimedia Enterprise, in an interview with Reuters. He added that the organisation had refined its commercial offering to encourage companies to move from free scraping to paid access.

The partnerships aim to create a more sustainable funding model for Wikipedia, whose primary income still comes from small public donations. The platform’s content is written, edited and fact-checked by about 250,000 volunteer contributors worldwide.

Microsoft said the agreement reflects a broader effort to support reliable information sources as AI systems become more widespread. “Access to high-quality, trustworthy information is at the heart of how we think about the future of AI,” said Tim Frank, Microsoft’s corporate vice president, adding that the partnership would help ensure contributors are valued within the AI ecosystem.

The Wikimedia Foundation also announced leadership changes, naming former U.S. Ambassador to Chile Bernadette Meehan as its new chief executive, effective January 20, following her appointment reported by Reuters last month.

The post After Accusing AI Firms of Unauthorised Data Use, Wikipedia Strikes Partnerships With Microsoft, Meta and Amazon appeared first on Tech Digest News.

In a blog post, the Foundation said the product allows companies to use Wikipedia’s content at scale without straining its servers, while also supporting the nonprofit’s mission.

The statement follows a rise in AI bots scraping Wikipedia’s content, often disguising themselves as human users. The organization said AI-driven traffic had inflated visitor numbers earlier this year while genuine human traffic declined by 8%.

Wikipedia said it expects AI firms to credit the human contributors whose work powers generative AI systems and to promote opportunities for users to visit and support the encyclopedia directly.

“With fewer visits to Wikipedia, fewer volunteers may grow and enrich the content, and fewer donors may support this work,” the post stated.

Earlier this year, Wikimedia outlined its own AI strategy, emphasizing that it would use artificial intelligence to assist editors — not replace them.

The post Wikipedia Slams AI Firms for Unauthorised Scraping, Demands Fair Use appeared first on Tech Digest News.

Speaking at the CNBC Technology Executive Council Summit in New York on Tuesday, Wales said, “The LLMs he’s using to write it are going to make massive errors. We know ChatGPT and all the other LLMs are not good enough to write wiki entries.”

Musk launched Grokipedia last week as an AI-powered alternative to Wikipedia, claiming it would “exceed Wikipedia by several orders of magnitude in breadth, depth, and accuracy.” Wales, however, dismissed those claims, saying Musk misunderstands how reliable knowledge is built. “His complaints about Wikipedia being ‘woke’ are misplaced. We focus on mainstream, verifiable sources — not random conspiracy theories — and that’s what keeps Wikipedia credible,” he said.

Wales noted that Wikipedia spends about $175 million annually on technology, compared to the tens of billions of dollars tech giants are investing in AI. He also cautioned that AI research tools — including those from OpenAI, Perplexity, and Google — often hallucinate or generate false citations, which could be dangerous if used for academic or factual purposes.

He cited a recent case where a German researcher found fabricated references generated by ChatGPT, calling the situation “a mess.” Wales said the incident underscores the need for Wikipedia to “double down on neutrality and reliability” rather than chasing AI hype.

While acknowledging that Wikipedia is experimenting with limited AI tools to improve content discovery, Wales said developing its own LLM is not currently feasible due to high costs. He also confirmed that the Wikimedia Foundation is tightening its rules on how third parties access and use its data, after detecting bot traffic disguised as human users — much of it linked to AI training activities.

Wales concluded that Wikipedia’s mission remains unchanged: “We shouldn’t be ‘Wokepedia.’ We should be the internet’s most trusted neutral source — and that’s exactly what people expect from us.”

The post Wikipedia Founder Criticizes Elon Musk’s ‘Grokipedia,’ Warns of AI Errors appeared first on Tech Digest News.

Marshall Miller of the Wikimedia Foundation revealed in a blog post that improved bot detection revealed that much of the traffic spike seen between May and June came from bots designed to evade detection.

Miller linked the decline to search engines offering AI-generated answers directly, reducing referral traffic to Wikipedia, and to younger users increasingly turning to social video platforms for information.

While Wikipedia remains a critical knowledge source—its content is still widely used by AI systems—the decline raises concerns over volunteer engagement and donations.

Miller urged AI, search, and social media companies to direct more users back to Wikipedia and encouraged readers to support “trusted, human-curated knowledge.”

The post Wikipedia Sees 8% Drop in Human Traffic Amid AI and Social Media Shift appeared first on Tech Digest News.