It’s Open Season for ‘AI Slop.’ Here’s How to Survive It.

“Not all AI content is spam, but I think right now all spam is AI content.”

These were the words of Jonathan Gillham, founder and CEO of AI content detection tool Originality.ai, in an interview with The Register in April of last year. At the time, he and his team found that 10% of Google search results pointed to AI content. At the rate it’s going, experts fear that the trend could lead to a model collapse.

However, it seems that the wind has somewhat shifted as platforms introduce features and policy changes to curb low-quality AI content, more commonly known as “AI slop.” YouTube, for instance, recently updated its “repetitious content” rule to make it ineligible for monetization, urging content creators to produce original content. (1)

While there’s still a lot of work to be done, such measures are a good start. And you don’t want to be on the receiving end of the hunt for AI slop when full-blown open season starts.

The Dangers of Low-Effort AI Content

While I can’t say I’m a fan of late-night talk shows, I know well enough that their coverage of a topic warrants a serious discussion. As it happens, HBO’s Last Week Tonight with John Oliver ran an entire episode on AI slop last June. You can watch the 30-minute piece below.

If you can’t spare 30 minutes, Oliver’s take is that the proliferation of low-effort AI content poses a grave danger to Internet users. Not only because of the risk of being fooled, but it also risks eroding trust in traditionally-made content. As AI gets better at blurring the line between fact and fiction, he said that AI slop threatens the concept of objective reality.

For the record, blaming AI for the low-quality content running rampant across the Web isn’t entirely justified. The purpose of AI is to augment processes, not replace them—and online content creation is no different. Anyone who leaves the entire creative process to ChatGPT or any other AI tool is contributing to the spread of AI slop.

That said, it brings us to another problem. AI models rely on training data to improve their outcomes, which in this case is published content. With AI slop multiplying like cancer as we speak, there’s a good chance that future models will feed on such data.

The result is, as mentioned earlier, a model collapse. The slop created by the forerunner model becomes the training data for its successor, leading to a line of models that gets dumber with each iteration. Eventually, the model becomes useless.

It doesn’t help that, according to a computer scientist from the University of New South Wales in Sydney, chatbots are running on fumes. He predicts that all the good data will have been used up by 2032. He also disagrees with OpenAI CEO Sam Altman’s idea of using synthetic data as a solution, saying that’s not how chatbot models scale. (2)

Google’s Stance on AI Content

Amid some platforms’ actions against AI content, Google’s stance somewhat sends mixed signals. Its search guidelines state that using AI for creating content isn’t against the rules, as it can be helpful when used correctly. Additionally, AI content is treated the same way as non-AI content and is subject to the same guidelines. (3)

Some SEO experts don’t share the same sentiment, especially following the March 2024 Core Update. While the update didn’t specify AI content verbatim, an expert pointed out that it’s still being penalized because AI content can’t meet Google’s quality standards. In his piece published on Search Engine Journal, he cites the following as evidence: (4)

• Firsthand experience, as required by E-E-A-T
• Google News results favoring human authors
• Google Perspectives favoring human-made posts
• Author background and expertise requirement

Then, last April, Google introduced new updates to its Search Quality Rater Guidelines, a rulebook for its team of human search quality raters. At the heart is generative AI, being defined in the guidelines for the first time. Other updates include: (5)

• Reorganizing and expanding on spam definitions
• Guidelines for delivering Low and Lowest ratings
• Imposing Low ratings for “filler” content
• Targeting content with exaggerated or misleading claims

The cherry on top happened last May with the introduction of the SynthID Detector. As the term implies, this tool detects online media for a SynthID-Text watermark, which identifies the content as AI-generated. The addition of this watermark to the data doesn’t affect the training of large language models (LLMs). (6)

As of this writing, Google technically won’t penalize AI-generated content. That said, given everything that was just discussed, don’t expect it to put AI-generated content high up the results page (or mention it on AI Overviews).

Surviving the AI Slop Hunt

It may seem that the best way to avoid being sanctioned for AI slop is not to use AI to make content. The problem is that we’re already in too deep, and turning our backs on this piece of technology will do more harm than good in the long run.

Source:SurveyMonkey

As the numbers from a SurveyMonkey study in 2024 show, AI tools are all but a staple of various business tasks. They become more important in SEO and content creation, with half saying they’re dependent on such tools. Considering how search engines have been introducing AI-powered features over the past years, the reliance on AI is understandable.

Still, preventing your content from being branded as low effort will require plenty of work. The good news is that you’re probably doing some of it right now.

Current SEO Practices Still Work

At the recently concluded Search Central Live in Bangkok, Google search advocates Gary Illyes and Cherry Prommawin said there’s no need for a separate AISEO strategy yet. They stressed that Google’s AI features were built on the same foundation as legacy search, and that AI augments current systems (e.g., RankBrain) rather than replaces them. (7)

As such, following Google’s search guidelines is still in your best interest. In fact, adding value is the best way to avoid having your content—made with AI or otherwise—seen as slop. And the best way to do that is to adhere to E-E-A-T.

We’ve explained E-E-A-T, which stands for Experience, Expertise, Authoritativeness, and Trustworthiness, a fair bit here. While not a ranking factor, search quality raters employ E-E-A-T to help Google’s algorithm identify and favor “helpful, reliable, people-first content.”

All the more reason to follow E-E-A-T if you publish Your Money or Your Life (YMYL) content.  YMYL content contains information that can affect aspects of a person’s life, such as their financial stability or overall health.

Go Beyond Text-Based Content

One thing that I learned over my time in the SEO market is that articles aren’t the only form of content. If the majority of your content is text-heavy but isn’t ranking despite being well-written, it may be a sign to explore other media.

Videos are often the first thing that comes to mind. One of their main advantages is their ability to explain technical stuff through visuals, something that can result in more words to read when done via text. Also, they have their own way of ranking high in search results.

The same can be said for infographics and other forms of image-based content. As long as you don’t create content with AI with the intent to fool the algorithm or mislead users, your content will be fine.

Promote Outside Search Engines

Speaking of diversifying your content media, you can take this opportunity to promote your business in areas other than search engines. Google may be handling billions of searches daily, but it isn’t the only site they visit.

The industry even has a term for this: search everywhere optimization (sometimes called SEO 2.0). Whereas SEO is focused on search results, SEO 2.0 optimizes content for, well, basically wherever your target audience goes. I’ll save the brass tacks for when we explore the topic in-depth, but places outside of search include:

• Generative AI tools like chatbots
• Video hosting sites like YouTube
• Reddit and other online forums
• E-commerce sites like Amazon
• Social media platforms like Facebook
• Review aggregator sites
• App stores, if applicable

A solid presence on multiple platforms helps establish a smooth customer journey. Here’s a look at how various media can support the process.

The Takeaway

AI slop is multiplying on the Internet at an alarming rate, posing a risk to not just users but also the AI tools they’ve come to rely on. As search engines and other platforms put their foot down against its spread, businesses should step up their content creation. The world may have room for AI content but none for low-effort AI content.

References:

1. Response to creator questions about YPP policies (July 2025) – YouTube Community [Internet]. Google.com. 2025 [cited 2025 Jul 24]. Available from: https://support.google.com/youtube/thread/356734251
2.      AI will run out of data within a decade – then what? [Internet]. Cosmos. 2024 [cited 2025 Jul 24]. Available from: https://cosmosmagazine.com/technology/ai/ai-will-run-out-of-data/
3.      Google. Google Search’s guidance about AI-generated content | Google Search Central Blog [Internet]. Google for Developers. 2023. Available from: https://developers.google.com/search/blog/2023/02/google-search-and-ai-content
4.      Montti R. Google’s Algorithm Hates AI Content? How To Make Google Love It [Internet]. Search Engine Journal. 2024. Available from: https://www.searchenginejournal.com/googles-algorithm-hates-ai-content-how-to-make-google-love-it/510854/
5.      Goodwin D. Google quality raters now assess whether content is AI-generated [Internet]. Search Engine Land. 2025. Available from: https://searchengineland.com/google-quality-raters-content-ai-generated-454161
6.      Kohli P. SynthID Detector — a new portal to help identify AI-generated content [Internet]. Google. 2025. Available from: https://blog.google/technology/ai/google-synthid-ai-content-detector/
7.      Bastian M. Google says AI content is fine, and SEO basics still apply to AI-powered search [Internet]. THE DECODER. 2025 [cited 2025 Jul 25]. Available from: https://the-decoder.com/google-says-ai-content-is-fine-and-seo-basics-still-apply-to-ai-powered-search/