AI-Generated Content: Can Google Detect It?

If you're creating AI-generated content on your website to rank on Google and publishing it without being edited by a human author, don't be surprised if your rankings drop suddenly.

It is against Google guidelines to scrape texts, search results, and merge other web pages generated through automated tools.

Likewise, you don't have to worry about AI-assisted content production with a human author’s review and final editing.

It's not really the "source" of the content that is Google's concern - it's the "quality."

Most AI-powered content writers are only a couple of steps away from the old-fashioned "content spinners." Some newer ones are very similar to the game "Mad Words"... (the game where you give a noun, adjective, verb, etc., and then fill in the blanks for humorous results).

Part of the problem stems from the application of the approaches. Another issue is that people simply don't understand and just use the "code."

As for usage — it's a little different from scraping existing/competing pages, compiling words/phrases, mapping structure/flow, etc.

So long as real human amends and edits, no harm.
Just publishing the MGC is likely to hurt your SEO.

John Muller says that content made with AI falls into the category of auto-generated content, which could lead to a manual penalty.

I think that for now, Google’s algorithms may lack the ability to detect AI-generated content without the help of human reviewers.

People using AI tools to generate content are still breaking the Webmaster Guidelines, according to Google. If Google’s web spam team happens to find it, they’re authorized to take action on it.

Google has recently updated the definition of automatically generated content in Google Webmaster Guidelines to read as "automatically generated content with the intent to manipulate search result rankings.”

With MUM and similar algorithms becoming more and more decisive in rankings, I think Google will be getting much better at detecting auto-generated content. That's why human proofreading is essential in this matter. 

I love all the information about MUM Multitask Uniform Model (MUM) update.

MUM is a multimodal algorithm designed to provide answers to complex queries by concurrently assessing information across multi-language text, images, video, and audio.

We should monitor how this update works in practice. 

Most likely there will be some flexible amendments and the algorithms. Google can't forbid all kinds of AI content, as it's a part of technological progress.

Before posting, we should check the content for SPAM.

From search engine points of view, content quality is more important than the content production method.

It might be an early signal that Google will be in the NLG market in the future. Google Docs already has some hints for that. They use generated text for KPs.

AI content isn’t the end-all-be-all for creating valuable content. Those insights lie outside of AI’s neural network.

Tool to get a rough draft started or expedite your research? Surely it will be low-value content just for keyword rankings, fighting a losing battle.

• It shouldn't make the blindest bit of difference.
• It's not like it's a free pass, or permission to produce garbage.
• Likewise, it's just a clearer explanation of their stance/perspective/preference.

Don't publish garbage for bots.

AI content is just as bad as Gibberish content to people from Google.

Google's entire business model can collapse if spam is not duly kept under control.

The availability of affordable NLG-based solutions is a potential threat for Google and for the Web in general. SEO and the content industry need to be ethical and user-centered to succeed.

There is an article titled "What site owners should know about Google's core updates,” Google directs site owners who experience organic traffic drops after core algorithm updates.

Unfortunately, the answers to questions such as the quality of the content in this document, whether it is based on actual data, factual errors, and whether the title is exaggerated cannot be given by the classical Google algorithms we know.

However, an artificial intelligence-supported algorithm such as MUM can easily do this. Therefore, unless finalized by a human author, Google can easily detect such content, and the site's rankings may drop.

Looks like it’s a big secret. I think it all comes down to auto-checking grammar, spelling, and behavioral factors. Similar to checking emails.

Ultimately, due diligence comes down to human responsibility. And until an AI can actually be held accountable for its mistakes, Google’s doing everyone a favor by marking anything the machines have to say as spam.

AI-Generated Content Detectors are getting better, but if the Language Model is unique, they need a human observer.

I may find my faults with Google, but the past decade has shown a Massive improvement in catching garbage. But, it's not AI content they want to catch, it's crap content they want to filter.

So long as the content produced is deemed as "quality", and doesn't trip any of the existing spam flags, it will likely pass. Then it seems to be a matter of whether it passes the SERPs (users, links, etc.) — not so easy to pass.

I think Google likely does a great job of understanding language in something written with statistical models.

If Google AI can be trained to understand the quality of content and whether it meets user intent, it is 100% possible to train an AI content creator to meet Google's expectations.

I think that in a few years, these tools will improve significantly, and Google may have to reevaluate its approach.

Most likely, it is a subject for Google to test. The user reaction is the most significant criterion.

It may turn out that the user was just looking for such text that was automatically generated :)

AI will present new opportunities and capabilities to improve the content experience.

While it is possible for a society to behave irrationally and choose to use it to their detriment, I see Google will bring new updates on how to use such tech in a more likely outcome.

NLP has been looking at and using MGC for decades and has done fairly well in certain cases (these being topic-specific, data-driven, heavily trained, etc.).

The major change is the size, scope of models, and the availability (or easier implementation) of scripts that access those models.

Whereas before you had to know the data structures, and/or linguistics, or implement systems that you had to train, etc., now you can just connect to an API with parameters.

Google doesn't really care if it's MGC, human-made, hybrid, or produced by blind mice singing psalms — so long as it's not garbage, they'll take it :D

At the end of the day, it's all about the quality of the brief / of the prompts.

You can use AI to accelerate the writing process, but you have to give the damn machine a proper brief, otherwise, you end up with irrelevant nonsense.

Before using these tools, I think it is necessary to understand the concept of Semantic SEO very well. Then, it is essential to read the articles and watch the videos of experts such as Bill Slawski, Koray Tuğberk GÜBÜR, and Marie Haynes.

Then you can use one of the artificial intelligence-supported content production tools.

Google Docs has started providing AI-powered content summaries. I recommend that you edit the content you create in artificial intelligence-supported content production tools in Google Docs to make sure that the content summaries are created correctly by the algorithm.

So you can see if Google's AI-powered algorithm correctly understands your content. If not, you can rearrange the content.

For sure, it depends. For example:

• research a topic and make a rough draft with bits of text from sources;
• rewrite that with AI tools;
• check for engagement by a content maker & humanize the text;
• check for correctness, clarity and uniqueness with the AI tool;
• proofread it by a responsible person.

1. Collecting data.
2. Training a unique language model.
3. Generating text.

However, using existing technologies is just being the first candidate to be out-ranked in the future.

It all depends on goals.

• Some people want a system that aids in guiding the writers (the "brief" machines listing words for inclusion, etc.)
• Others want something that actually produces content (the "writer" machines outputting via templates/chunks).

The process/approaches you take depend on the tool and your expectations.

As it currently stands — you still need a fair bit of human intervention. And as tools (makers) mature, and the users get a better grasp of how to work, things will change.

Generate article → read article → use Google NLP to find out missing entities → add value (image, table, video) → edit H1-H2-H3 → correct grammar → add an outgoing internal link with relevant anchor and publish it.

I believe AI-generated content will replace the content writer. It will not replace the copywriter in the short term.

It's a harder skill to teach an AI tool, but when it comes to AI, you should never say never; better say “it depends” :)

Will AI take content marketers’ and copywriters’ jobs? Never. AI is far from perfect.

Still, it is no doubt gaining a foothold in the content production process. Those marketers who do not experiment and discover the potential of AI for content could be left behind.

Simply, yes. Eventually, it will happen. :)

But, this is a general outcome of the future. AI can replace politicians, or judges as well.

Yes and No.

• Yes: scripts will be able to produce common, everyday content — so long as there are plenty of existing examples and data.
• No: without data/examples — they won't be able to produce new insights, etc.

At the end of the day — the machines are much like us; they are reliant upon experience, understanding, information, and resources. If those things are lacking, the output will be too.

I don't believe that AI will replace human writers. Humans are too desperate :)

Here's what an AI has to say about it :)