The black hat SEO resurgence – all thanks to LLMs

Are black hat SEO tactics making a comeback? Find out why your brand should avoid the risk.

By 

, 

published on 

Unless your business is operating in an industry where a churn-and-burn strategy isn’t a problem, black hat SEO was starting to look like a dying art. 

As SEO agencies pivoted to more ethical ways of ranking websites, through robust content strategies, time-consuming (and often expensive) research backed by outreach and digital PR, and quality-over-quantity-based linkbuilding strategies, all bolstered by a more rounded marketing strategy, it looked like those using “bad” SEO tactics days were numbered. 

But then Large Language Models (LLMs) like ChatGPT and Google’s Gemini came along, and inadvertently opened a back door to gaming the system once again.

As search engines shift toward AI Overviews (AIOs) and AI tools gain in popularity, a new breed of LLM Optimisation is emerging, and it’s looking a lot like the Wild West of the early 2000s.

We are currently witnessing a massive resurgence in manipulative tactics, rebranded for the AI era as Generative Engine Optimisation (GEO). Here is how the new wave of black hatters is gaming the machines, and why it’s a dangerous game for any reputable brand to play.

A brief history of black hat SEO

Historically, black hat SEO was about tricking a crawler. Tactics like cloaking (showing different content to Google than to users), keyword stuffing, or Private Blog Networks (PBNs) were designed to inflate a site’s authority artificially.

Eventually, Google got smart. It stopped looking just at the quantity of signals and started looking at the quality and intent. For a decade, “white hat” digital marketing became the only viable path for sustainable growth. 

However, LLMs can be easily manipulated if fed the ‘right’ signals, creating a number of loopholes that black hats are now exploiting with glee.

How are LLMs being manipulated?

LLMs do not know things; they output responses based on the data they were trained on or the data they fetch via Retrieval-Augmented Generation (RAG). Black hats are now focusing on ‘