These Days, AI Keeps Rewriting the Marketing Rulebook

Lately, my feed has been blowing up with AI news again.

Friends in marketing have been forwarding it wave after wave. Some ask, "OpenAI is putting ads in the free tier — can we still freeload?" Others share, "Apple just handed Siri to Google." And one buddy cut right to the chase: "The boss is making the team do AEO — what the heck is AEO, anyway?"

I went back through all of it. Here's my bottom line:

AI has stopped being about "thinking smarter for you" and started being about "taking money out of your pocket more smoothly."

Why do I say that? Let me walk you through three fronts.

---

One: The Entry Point Is Starting to Charge You

What does "charging at the entry point" mean? It means the AI you use is starting to figure out how to make money off you.

OpenAI says it will test ads in ChatGPT's free tier and its Go plan. How much is Go? $8 a month. Do the math — that's the price of a cup of coffee. What you get is an assistant that isn't quite free anymore, but still cheap.

If you work in marketing, this should make you pause. How did search get started? It drew everyone in with free, then sold them next door to advertisers. Now OpenAI has quietly wrapped that old playbook in AI.

It did add a caveat: ads won't affect answers — they'll be kept strictly separate and clearly labeled. Whether you believe that is up to you. But the direction is unmistakable: the free-lunch era has a countdown in AI, too.

Apple's move that week was just as heavy. It announced that its own foundation models would run on Google's Gemini and Google Cloud. In other words, Siri is no longer the clumsy bird Apple built behind closed doors — it borrowed a rival's hands and learned to talk again.

When the giants start lending each other ladders, the smart move isn't to sit back and watch the show — it's to read which way the wind is blowing.

And Anthropic went even harder. It rolled out a thing called Cowork with a single message: "Make Claude usable by everyone who isn't a coder." What does that mean? Using Claude used to be a programmer's privilege. Now everyday people are lining up too.

That gave me chills. The AI entry point is shifting from "search once before you leave" to "living at your fingertips every day." Once that entry point fills up with people, that monetization gold mine isn't going to lack miners, right?

Two: AI Is Rewriting the Rules of Search

The second front is where the traffic flows.

You've probably noticed: it used to be "open Google, search, click through ten links." Now the answer is written right in the dialog box, and links are just footnotes. After the search, whose traffic wins is up to the machine.

What is AEO? AEO stands for Answer Engine Optimization. For the same question, when you ask GPT and Gemini separately, does the AI happen to mention you while it assembles its answer?

What is GEO? Generative Engine Optimization. Same logic: it's about "writing yourself into the generated answers."

Sounds like something from the humanities. But the recent moves show it's gone from concept to a real, serious business.

Google struck first. Its Trends Explore page got an upgraded side panel that uses Gemini to automatically compare related trends for you, and it even hands you a few "dig deeper" prompts. What's the point? It's doing subtraction: you no longer need to understand how to search — AI has done the searching for you.

Hot on its heels is "Personal Intelligence," another test feature from Gemini. What does it do? It digs through your Gmail messages and the photos and videos in your gallery, then comes back chatting like a thoughtful little assistant. Your personal data is being used to tune its answers. Think about it for a second — it's kind of neat, and also kind of spine-chilling.

Microsoft hasn't been idle either. It just published a PDF guide teaching you how to feed and organize your data now that AI search, AI assistants, and AI browsers are here. The official playbook is turning "how to be seen by AI" into a public course. That signal carries serious weight.

Even the question "will AI quietly raise prices?" got an official response. Google says the personalization in its AI shopping mode is about "recommending better options to you," not "quietly marking up the list price." Do you believe it? I believe half of it. But "AI wants to stick its nose into your shopping decisions" — that part is here to stay.

Three: The Numbers Are on the Table — Don't Be the Last to Move

Listening to insiders talk isn't enough. The data is what proves it's real.

There's a survey of 250-plus digital-marketing decision-makers, and the results: 97% say doing AEO paid off in 2025, and 94% plan to double down in 2026.

97%. 94%.

Let me check my own math. Yeah. The "I won't touch AI search" crowd is no longer the mainstream. If you've been treating this like a sideshow, pay attention — the game has moved into its second half, where you count the returns.

Beyond the big tech moves, there's also a batch of handy tools that have shown up recently. Let me pick a few for you.

One is about seeing traffic structure. Ahrefs released an "AI vs search traffic" analysis tool that lets you separate traffic from traditional search and traffic coming through AI platforms. Which channel is quietly growing, which is quietly shrinking — one chart shows it all.

One is about testing "what AI sees when it looks at you." There's a platform called Columbus that specializes in AI engine optimization. It lets you check: when people chat in ChatGPT, Claude, Gemini, or Perplexity, does your brand get named? If it gets named, you've won the entry point.

And on a side note, OpenAI also released "ChatGPT Translate": it supports 50-plus languages, and you can type, speak, or snap a photo — it works out the "tone and context" before translating. Before you send copy to multilingual clients, run it through once. It'll help you hit the right note and avoid embarrassing blunders.

Finally, there's a question nobody can dodge: attribution. Clicks are falling, so where does the conversion credit go? The answer people are feeling their way toward: first, convince your boss that "fewer clicks isn't a bad thing"; then use metrics like "impressions" and "mentions" as stand-ins; and finally, find a way to link exposure to revenue. Nobody has fully cracked this path, but someone has to take the first step.

Closing: Follow Along — or Wait to Be Called Out?

Back to the friend from the opening.

I eventually gave him the most plain-spoken advice I had: if you want to understand AEO, go search your own name and see whether the AI-generated answers mention you.

If you're not mentioned, fill the gap. If you are, let it ride.

AI has never been waiting at the door for you. It keeps moving. If you stay where you are, it will naturally pass other people's houses first.

I'm not going to feed you that "adapt or die" line — it isn't true. Here's what I actually feel: the new rules are just beginning to keep score, and the one who moves first — even if they don't profit — at least doesn't pay the tuition.

I hope you never have to say the words "if only I'd known."