Thanks to AI? Links to my site started arriving from an unexpected route

Upon reviewing the Links report in Search Console for a site I maintain, an unusual pattern emerged. While the majority of external links directed to my home page, originating from established sources such as Hatena and Yahoo! Japan, the links pointing to my blog posts were unique. These were contained within the "References" section of articles penned by authors I had never encountered before.

To gain further insight, I meticulously traced each linking page, examining the nature of articles linking to my site. In Search Console's Links report, I proceeded as follows:

1. Listed all pages with external links, totaling 98 pages.

2. Focused specifically on the 68 blog posts, reviewing the linking sites one by one.

3. For those links originating from note, a prominent Japanese blogging platform, I delved deeper, extracting the URLs of each linking article through the note's public API.

From this analysis, it became clear that 68 blog posts had acquired external links, amounting to 125 in total. These links were sourced from 36 distinct sites, with notable contributors including note (19 posts), Hatena Blog (5 posts), Yahoo! Japan (3 posts), and 33 other sites (ranging from 1 to 6 posts each). The bulk of these links to my blog posts predominantly emanated from note and personal blogs.

The note links traced back to 31 articles, revealing a distinct format - numbering each citation with the site name, date, and title, mirroring the structure often seen in academic publications.

Interestingly, some of these 2026 articles included a "References" or "Sources" section, explicitly listing my posts, often accompanied by markers like [1, 2] within the main text. This meticulous citation style was notably uncommon in personal notes. Conversely, AI-generated research outputs frequently adopt this structured approach.

While I cannot definitively ascertain whether these articles were authored by AI, the similarity in their format is striking. Notably, one author cited my site across eight different articles, underscoring the breadth of their referencing.

Upon scrutinizing by month, a notable surge was observed in July 2026, coinciding with an increase in citations. This uptick was traced to a company-owned media site within the IT sector, which inserted references automatically through a system that identified related external articles and appended them at the end of each piece.

Tagged as "source: injected", these references appeared uniformly across various related sites, highlighting a systematic approach to link-building. Upon closer inspection, I confirmed that the site's links were marked as "noFollow", a fact shared by the 11 reference cards on the company site, despite the absence of any discernible increase in traffic or search visibility for my site.

To gauge the impact of these links on my site's search performance, I juxtaposed Search Console data from the 21 days preceding and following each post's citation. Despite the presence of 19 blog posts cited from note articles, the results were inconclusive. Impressions for these posts fluctuated slightly, with some experiencing a modest decline, while others remained stagnant.

Importantly, none of the cited posts registered a significant rise in search rankings, suggesting that while the citations were mechanically recorded, they did not translate into enhanced search performance. This finding aligns with the noFollow attribute applied to these links, indicating that search engines do not assign direct ranking value to such links.

What emerged most compellingly was the observation that the 19 blog posts, already well-positioned within search results, bore citations in subsequent articles. This phenomenon suggests that established high-ranking content may garner attention from AI-driven research, irrespective of citation-induced search boosts. The overarching implication is that while AI-generated citations may not directly enhance search rankings, they signify a broader trend towards automated information sourcing and dissemination.

As such, monitoring and understanding these patterns could offer insights into future content strategies and the evolving landscape of online visibility.