There’s growing debate about whether the discussion around LLMs.txt has become a cycle of confusion, raising questions about what it really means for AI-driven search visibility.
A number of well-known WordPress SEO plugins, as well as other content management platforms, now include the option to create an LLMs.txt file. This has led many to believe that adopting it could be the next big step in search optimisation for AI platforms.
However, the reality is rather different. At present, LLMs.txt is nothing more than a proposed standard, and no major AI platform has officially implemented or recognised it.
This begs the question: why are so many businesses quick to jump on board with a feature that has no practical application yet?
For some software providers, the answer is fairly simple — their users are asking for it. If customers believe it’s important, even without evidence of its value, developers feel pressure to provide the functionality.
On the other hand, many website owners and marketers feel they should adopt LLMs.txt simply because their preferred tools now make it available. In turn, this reinforces the idea that it must be essential for SEO, even though its usefulness has not been proven.
A recent discussion on Reddit highlighted this very issue. Contributors pointed out that much of the hype around LLMs.txt stems from speculation, rather than actual evidence that it improves rankings or visibility in AI-powered search tools.
The result is what some describe as a self-reinforcing loop: tools create the feature because users expect it, and users adopt it because the tools make it available — even though AI platforms aren’t actually using it.
Third Party SEO Tool And LLMs.txt
Google’s John Mueller recently responded to confusion surrounding LLMs.txt in a Reddit thread. The question came from someone whose SEO tool flagged the file as missing, returning a 404 error. This gave the user the impression that the file was required.
The user asked:
“Why is SEMrush saying /llm.txt is a 404? I know I don’t have it on my site, but I’ve heard it’s pointless and unnecessary. Is that right?
If I do need it, how should I create one?
Thanks.”
It seems the person was unsure because the SEMrush audit appeared to suggest the file was necessary. While it’s not clear what their exact audit results showed, SEMrush’s own documentation states that without a proper LLMs.txt file, websites could risk being misinterpreted by AI systems.
According to their notes:
“If your site does not have a clear llms.txt file, it could risk being misrepresented by AI systems.
…This new check makes it easier to identify potential issues that might limit your visibility in AI search results.”
Although the wording suggests a “risk,” the reality is that there is no actual danger. At present, no AI platform makes use of the LLMs.txt file. This likely explains why the Redditor followed up with the question: “If I need it, how do I build it?”
LLMs.txt Is Unnecessary
Google’s John Mueller has clarified that the LLMs.txt file is not needed.
In his response, he noted:
“Nice observation! In SEO, it’s always crucial to spot misleading or incorrect advice early on, before wasting time on something that serves no real purpose. Always approach things with a questioning mindset.”
Why AI Platforms May Choose To Not Use LLMs.txt
While John Mueller has made several casual remarks about the lack of value in LLMs.txt, there don’t appear to be any official statements from AI platforms explaining why they have chosen not to adopt it or the related .md markdown files. That said, there are clear reasons why such platforms would steer clear.
One of the strongest arguments against LLMs.txt is that it cannot be relied upon. Standard on-page content is generally more trustworthy because both users and AI systems see the same thing.
With an LLMs.txt or markdown file, however, the situation is different. There is potential for manipulation, as some SEOs could insert details that aren’t present in the visible HTML content to gain an advantage. Assuming that publishers wouldn’t attempt to exploit this is unrealistic.
There are already examples of dishonest SEO practices, such as hiding text or embedding AI prompts within HTML in an attempt to mislead. Research published in 2024, titled Adversarial Search Engine Optimisation for Large Language Models, demonstrated that manipulating LLMs is possible through a technique called Preference Manipulation Attacks.
According to the study:
“…an attacker can trick an LLM into promoting their content over competitors. Preference Manipulation Attacks are a new threat that combines elements from prompt injection attacks… Search Engine Optimisation (SEO)… and LLM ‘persuasion.’
We demonstrate the effectiveness of Preference Manipulation Attacks on production LLM search engines (Bing and Perplexity) and plugin APIs (for GPT-4 and Claude). Our attacks are black-box, stealthy, and reliably manipulate the LLM to promote the attacker’s content. For example, when asking Bing to search for a camera to recommend, a Preference Manipulation Attack makes the targeted camera 2.5× more likely to be recommended by the LLM.”
This shows that whenever there is a loophole, someone will eventually try to take advantage of it. A separate file designed purely for AI systems could easily be misused as a tool for spamming.
Because of this, depending on on-page content is far more secure than creating additional markdown files that can be crafted solely to influence AI. Ultimately, this is why many see LLMs.txt as inherently unreliable.
What SEO Plugins Say About LLMs.txt
The team behind the Squirrly SEO plugin for WordPress have openly admitted that they added support for LLMs.txt purely because their users requested it, rather than because it has any proven benefit. They also made it clear that the feature has no effect on how visible content is within AI-driven search platforms.
In their own words, they explained that many users wanted the option included, so they decided to deliver it. However, they also issued a clear warning, reminding users that LLMs.txt will not make their sites suddenly appear in AI search results. At present, there is no evidence at all that it influences how AI systems promote or display content.
This approach is refreshingly honest, as it balances meeting customer demand with the responsibility of pointing out that the tool itself is unnecessary for search visibility.
On the other hand, Rank Math has taken a different stance. While Squirrly has downplayed the usefulness of LLMs.txt, Rank Math has suggested that AI chatbots actively rely on the content structured in markdown files when generating answers.
To be fair, Rank Math is correct in its explanation of what LLMs.txt is designed to do and how it functions. However, the issue lies in overstating its value. Their message implies that chatbots draw directly from curated markdown versions of content, which simply isn’t the case.
They state that by providing this curated content, websites increase their likelihood of being cited correctly, represented accurately, and more easily found within AI-powered search results.
The reality, however, is very different. Current AI chatbots do not pull from markdown files, nor do they use structured data in this way. Instead, they continue to process the standard HTML content presented on websites.
This makes Rank Math’s claims misleading, as they overemphasise the role of LLMs.txt when, in truth, it has no practical influence on AI search behaviour.
Yoast SEO, meanwhile, adopts a more cautious position, sitting somewhere between the stances of Squirrly and Rank Math. Their explanation outlines what LLMs.txt is intended for, but avoids exaggerating its benefits by using softer language such as “can” or “could.”
This middle-ground approach provides users with a fair understanding of the concept without promising results that simply aren’t there.
In comparison, Squirrly’s transparency feels more credible: they added the feature because people asked for it, but they don’t mislead users into believing it will transform their search rankings. For now, that honesty may be the most helpful stance of all.
The LLMs.txt Misinformation Loop
The discussion around LLMs.txt has turned into something of a feedback loop. Business owners and SEO professionals, worried about how their sites might appear in AI-driven searches, often feel pressured to take action. For many, adopting LLMs.txt has become that “something” they can do.
SEO software providers, in turn, feel obliged to include LLMs.txt as a feature in their tools. This reinforces the impression that it is essential, even though it isn’t, and helps fuel the ongoing cycle of confusion.
The irony is that the growing concern about AI visibility has pushed many towards using LLMs.txt, despite the fact that it remains only a suggested standard. At present, no AI platform has implemented or relied on it.
More Digital Marketing BLOGS here:
Local SEO 2024 – How To Get More Local Business Calls
3 Strategies To Grow Your Business
Is Google Effective for Lead Generation?
How To Get More Customers On Facebook Without Spending Money
How Do I Get Clients Fast On Facebook?
How Do You Use Retargeting In Marketing?
How To Get Clients From Facebook Groups