AI is transforming search engine optimisation, but the rush to automate content production risks reviving some of the industry’s worst habits. The challenge isn’t whether we should use AI. It’s whether we’re prepared to retain the human judgement that makes content valuable.

We’ve seen this behaviour before

I’ve worked in SEO since 2014. While the industry had already moved beyond some of its earliest excesses by then, the legacy of keyword stuffing, article spinning, doorway pages and artificial link schemes was unmistakable. I have watched search engines, businesses and practitioners continue to grapple with the consequences of trying to manufacture ranking signals rather than genuinely useful information.

Some of those older techniques worked, at least temporarily. But they also contributed to an internet cluttered with repetitive and low-quality material. Search engines responded with increasingly sophisticated systems intended to identify manipulation and surface more helpful results.

Now generative AI has introduced extraordinary capabilities, while making the production of large quantities of convincing-looking content exceptionally cheap. Are we about to repeat the mistakes of the past, only this time on a much greater scale?

The technology has changed. The temptation hasn’t.

AI offers enormous opportunities for SEO professionals. It can assist with analysing complex datasets, identifying technical problems, researching subjects, organising information and examining search intent. Used thoughtfully, those capabilities can improve both the quality and efficiency of our work.

The concern arises when productivity becomes the objective rather than the quality of the result.

A business that might once have commissioned ten carefully researched articles addressing customer questions can now generate hundreds of pages targeting combinations of services, suburbs, questions and keyword variations. But producing content is not the same as producing knowledge.

A thousand automatically generated articles may look impressive in a content management system. That does not mean they provide a thousand pieces of useful information. Many may simply rearrange the same claims into different formats. The old assumption returns: if one optimised page works, surely producing hundreds will work better.

The problem isn’t the technology. It’s the assumption that volume creates value.

AEO risks becoming the next shortcut

Answer Engine Optimisation (AEO) and Generative Engine Optimisation (GEO) raise legitimate questions about how websites are discovered, understood and referenced by AI-powered search systems. Businesses naturally want their information to be accessible, accurate and well structured.

But AEO must not become another excuse to manipulate visibility. Extensive FAQ libraries, conversational question pages and structured content aren’t inherently problematic. A useful FAQ responding to a genuine customer need can be excellent content.

The distinction lies in why the material exists and what it contributes. If hundreds of variations simply reproduce information already available elsewhere, they add little. Automated production can also introduce errors, outdated claims and superficial explanations that sound authoritative without being adequately substantiated.

Human oversight means more than proofreading

It is easy to generate an article and ask someone to check spelling, grammar and readability before clicking Publish. That alone is not meaningful editorial oversight.

A responsible editorial process begins earlier: deciding whether the subject deserves coverage, understanding the reader, identifying missing information and establishing which sources and expertise are needed.

During production, someone must challenge assumptions, verify claims, test relevance and recognise oversimplification. Before publication, someone must determine whether the result offers anything genuinely useful.

Sometimes the most valuable editorial decision is that another page doesn’t need to exist. AI can support this process, but expertise and accountability should remain human responsibilities.

Google has already identified the problem

Google’s spam policies explicitly identify scaled content abuse: producing large amounts of content primarily to manipulate search rankings rather than help users. The policy is not restricted to AI-generated material.

Likewise, Google’s guidance on generative AI content recognises legitimate uses while warning against producing large volumes without adding value. The important distinction is purpose and usefulness, not whether a human or a machine typed the first draft.

The web needs original knowledge, not endless versions of the same answer

AI-powered search relies on trustworthy source material. If more of the web consists of machine-generated summaries of existing summaries, we risk an information environment in which content multiplies faster than genuine knowledge.

A technician documenting a difficult repair, a business owner describing an operational challenge or an SEO consultant analysing actual campaign findings can contribute experience not found in generic summaries. AI can help interpret and communicate that experience, but it should not be mistaken for the experience itself.

As generated information becomes more abundant, firsthand expertise, original research and independently verified evidence may become more valuable, not less.

The opportunity for experienced SEO professionals

I remain optimistic about AI in SEO. Being able to inspect larger datasets, investigate technical issues more deeply and spend less time on repetitive processes is genuinely valuable. It creates more room for strategy, analysis and quality control.

Using AI to identify content gaps is useful; automatically generating a page for every gap without checking whether it deserves publication is a different proposition. AI can assist with structured data, but wrapping an inaccurate assertion in schema markup does not make it true.

The strongest applications combine automation with subject knowledge, evidence and considered decision-making. That’s where I believe AI offers its greatest potential.

We should remember what SEO has already learned

Every major technology creates legitimate opportunities alongside the temptation to exploit perceived weaknesses in search systems. Generative AI is no exception.

The future of SEO and AEO should not be a competition to publish the most machine-generated content. It should be about making genuinely useful, accurate and original information easier to discover, understand and trust.

AI should amplify human expertise, not manufacture the appearance of it. If we preserve that distinction, AI can make SEO significantly better. If we don’t, we may revisit some of the industry’s oldest mistakes — only at unprecedented speed.

Is AI Taking SEO Back to the Dark Ages? Why Human Expertise Matters More Than Ever