An earlier post A further examination of AI in legal blogging examined recent changes in the use of Artificial Intelligence in relation to material used on the blog. It focussed on the output of L&RUK, our use of AI in relation to material obtained from various sources, and the potential impact of ChatGBT and Copilot. This present post examines referrals (i.e. effectively the input to the blog) using the statistical information provided by WordPress from various metrics: page views, visitors and referral bodies (websites, AI &c), and their geographical and time-related origin.
Background
WordPress provides its users with information on site traffic through a number of pre-set options[1] including: “Last 7 Days”; “Last 30 Days”; “Last 12 Months”; “Last 3 Years”. It also has details for: “Today”; “Month to date”; and “Year to Date”, which are useful in tracking the development of specific issues, but are of less value for comparative purposes. With L&RUK’s worldwide readership, the “Today” results are influenced by the time zone and different responses over the 24 hour period post-publication.
Likewise ” *** to date” information is of limited value when seeking comparisons on due to possible variations in the length of the time period chosen. Consequently, this present analysis uses information for the “Last 7 Days”; “Last 30 Days”; “Last 12 Months”; and “Last 3 Years”.
Referrals
“Referrals” are made to the blog from two distinct sources: from search engines[2] and through other web links[3]. Search engines are associated with 26.8% to 37.8% of the total number of page views, and of these, the majority are from either Google, (82.34% to 86.28%) with a significant, but smaller number from Bing, (10.4% to 14.0%). Of the remaining referrals through other links, the majority are either from X, or Facebook, with very few via ChatGPT[4].
There are also ~1,160 subscribers who receive email notification of each new post as it is published, although these are not identified separately.
Location
In the context of the blog’s strapline “Issues of law and religion in the United Kingdom – with occasional forays further afield”, it is hardly surprising thar the majority of readers are from the UK. There were 79.5% over the past three years, falling to 68.9% over the 12 months period when there was a significant readership in the the United States (17.1%) and China (12.5%). Recent UK readership is 71.41% (30d) and 71.69% (7d)[5].
Comment
Search Engines play a significant part in identifying L&RUK to potential users, and as noted earlier, they now rely on AI at almost every stage of the search process—from understanding the query, to ranking results, to generating summaries. Thus whilst the general trend of an increased use of AI is not reflected directly in the above analysis, there is likely to be an indirect effect though the Search Engine searches.
[1] A bespoke option is also available.
[2] Google; Bing; Duck Duck Go; Yahoo; ecosia; and others.
[3] X; Gmail; ChatGPT; Facebook; Inner Temple; Thinking Anglicans; Beaker Folk of Husborne Crawley; Edgepilot; Notebook; and others.
[4] X, 1.05% to 6.46%; Gmail, 0.91% to 1.87%; ChatGPT, 0.28% to 1.1%; and Facebook, 0.7% to 3.28%.
[5] However, comparing the subject matter of “top ten” posts for this 12 month period, there appeared to be little that would be of specific interest on China or the US.
Update: 4 April 2026 at 14:22. See Comments, below, on information from Copilot: viz. “The spike in 2026 Chinese page‑reads is almost certainly caused by AI‑related automated traffic—crawlers, LLM training systems, or AI‑driven referrals—rather than a sudden surge of human readers in China”.
This gave an interesting insight into one more of the many ways that AI is exercising steadily increasing impact in our lives, whether we are aware of it or (often) not. I was particularly intrigued to see all that interest coming from China –– As you comment, most of L&RUK’s topics are very specific to the UK, or more broadly, Europe; we might have expected more European response. So is the Chinese interest perhaps largely an accidental side effect of AI in the search engines?
I would agree with you both that the Chinese interest was perhaps largely an accidental side effect of AI in the search engines. Looking at the page reads for China for different time periods:
Last 7 Days, 0.96%; 30 Days, 2.78%; 90 days, 28.59%; 365 days, 18.03%; “All time” (i.e. since 2012), 1.16%.
The peak in China data was only experienced this year, and is currently back to at an expectedly low level.
The answer I received from AI Copilot was:
“The spike in 2026 Chinese page‑reads is almost certainly caused by AI‑related automated traffic—crawlers, LLM training systems, or AI‑driven referrals—rather than a sudden surge of human readers in China.
This conclusion aligns with:
• L&RUK’s own April 2026 analysis of AI referrals,
• The improbability of large‑scale human access to foreign religion‑law blogs from within China,
• The timing of UK parliamentary debates prompting AI systems to gather related material”
DavidP
That would certainly be my guess. And since the blog began, Akismet has blocked almost 700,000 login attempts – almost all, i would guess, from pornbots.