Today’s internet is on track to go the way of cassette tapes and DVDs – a distant memory for those of us who used it, but completely foreign to the next generation.
Big call? Not really. By now, we should be used to things being thrown out the window as technology evolves. And some of the updates we’ve seen from Google and OpenAI over the past month have shown us that big change is coming sooner than we may have anticipated. So, let’s break it down.
The players: Google vs. OpenAI
Google’s the established empire, defending its massive territory. OpenAI? They’re the rebellion. But OpenAI isn’t just trying to storm the castle, they’re building a whole new world from scratch.
The outcome will have consequences for both how we (as consumers) get our information, and who controls it.
What we’ve seen over the past few months is Google’s game plan has two parts – defend the empire and police quality for the whole internet.
Defending the Google empire
First steps? Plugging Gemini into literally everything they own. They pushed out the biggest AI upgrade Chrome has ever seen and launched Search Live in the US.
Chrome’s update also sees AI Mode embedded directly into the address/search bar for most of the world. We’re still yet to see AI Mode here in Australia and there’s no whispers about why it’s been delayed or when that’s coming to our shores. My suspicion? The News Media Bargaining Code. Because how can AI Mode be effective if it must be turned off every time someone asks for details of well, pretty much anything the Australian media might cover? Of course, that’s just speculation – I don’t know the real facts.
Cleaning up the internet
Their second recent move is through a huge spam update aimed primarily at useless AI content. To be clear, they’re not shutting down AI content, they’re only focused on the value it provides searchers.
Content that shows real human expertise, that’s original and trustworthy is rewarded. Content that just targets a keyword without meeting the searchers’ needs is deprioritised or even completely delisted.
What that means? It’s basically Google doubling down on what they’ve already said many times before. High quality, valuable content is key to visibility. Low effort, mass-produced content is the path to invisibility.
OpenAI
OpenAI on the other hand is focused on building a new playground that makes the old internet irrelevant. They’re building a new, self-contained experience, capturing people within their platform and giving them absolutely no incentive to leave.
A new world of e‑commerce
In the past month, OpenAI added Instant Checkout to ChatGPT in the US. That immediately remodels ChatGPT from a powerful discovery tool (and creating tool – but we’re just focused on search here) into a fully personalised e-commerce channel.
They’re starting with single item carts from Etsy sellers, using the new Agentic Commerce Protocol (ACP, co-developed with Stripe), but plan to expand to all Shopify merchants with multi-item carts (and likely multi-seller carts too) soon.
For now, products are ranked by relevance but expect that to change soon because…
OpenAI is building an ad platform
If you didn’t see this one coming, I’d like to live in your innocent life for a little while, please.
OpenAI is hiring to build their own ad platform. From my own personal experience as the lead content designer on a successful ad platform, I’d say be prepared to see this in beta by March 2026 at the latest. (I’d expect they’ve set a January deadline but there are always issues and delays…)
How will ChatGPT ads work? Entirely TBD but I’m predicting sidebar panels, sponsored messages integrated naturally within generated answers, and “upgrade” options for workflows and more complex queries.
Asking for recipe ideas? Expect ads that allow you to buy the ingredients and have them delivered within the hour, directly from Woolworths.
Looking for fashion ideas? Expect a line-up of styles from ASOS leading the way among organic suggestions.
Brainstorming a keyword strategy? SearchAtlas, Semrush and Ahrefs will be there showing how you can take greater control.
Or maybe not… we’ll see.
But is Google the free and open web we like to think it is?
OpenAI is building the infrastructure to meet most of your online needs within their platform. But Google has also been doing this for years and we all kinda pretended they weren’t. They have Google Shopping for commercial product queries. They prioritise YouTube over other video platforms. They have a vested interest in keeping you in their world – just like OpenAI does.
And with a 98%+ search ownership (or whatever the latest figures are), we came to subconsciously believe that Google IS the internet. It’s not. The internet is the open web; it’s free access to information without bias. And that’s not something we’ve seen since the late 1990s – well before most people were online.
Cloudflare’s pushing back
A couple of months ago, Cloudflare made the news for blocking AI crawling of sites by default. In September, they introduced the next step of that: a new Content Signals Policy to give website owners more control. The outcome is an update to robots.txt files.
For those who don’t know, a robots.txt file tells bots what to scrape and what to ignore. You could also use it to specify which crawlers are allowed and which aren’t – so, you could say Google can look everywhere but ChatGPT has to stay out of your archives if you wanted to. Most bots and crawlers obey the rules, some don’t… but the file today is pretty much a simple yes/no on content access.
Pre-AI, it was a good thing. It helped you rank on Google and other search engines.
Today’s zero-click search means it’s more complicated. Websites aren’t getting the traffic for their efforts, and it’s only going to get worse with OpenAI busy creating their walled garden.
With Cloudflare’s new Content Signals Policy, website owners get to determine what crawlers can do with their content after they’ve scraped it – whether they can list or rank it in a traditional search engine, use it for generative answers, or use it for training AI models.
Cloudflare is saying no to AI using their customers’ sites to train AI models by default. If enough high-value sites follow suit, the internet as we know it today won’t disappear quite so quickly.
So, what does this all mean?
Microsoft, Perplexity, Anthropic… and whoever comes next. To stay competitive, they’ll need to follow Google and OpenAI’s lead. We’ll each be locked into a platform eventually – but at least this time we’ll get to decide who we trust to deliver information.
1 Sistrix, ChatGPT Using Web Searches Less Frequently, September 2025
