Everything we build at The AI Collective is for YOU, our community. We're gearing up to launch more content, more virtual events, and more opportunities for our members to connect.

I want to hear from YOU directly. Please take 2 minutes to complete this quick survey for a chance to win $100 in Codex Credits. The more people who answer, the better we can build the community that YOU need.

-AJ Green, Executive Director

One safety story a week, pressure-tested for what's actually happening underneath the surface.

By Lindsay Gross

Ads arrived inside ChatGPT for free and Go users in the US on February 9, 2026, and in May OpenAI opened the self-serve platform that lets any business buy one. The reaction split between "this ruins it" and "what did you expect." An online product carrying ads is not new either. Instagram, TikTok, and Google have sold attention for years, ranking what you see to keep you engaged. The question worth asking is not whether a chatbot shows ads. It is what makes a chatbot ad different from the ones you already scroll past.

Designers have a name for the small choices that quietly steer you toward what a company wants instead of what you came for: dark patterns. The pre-checked box, the buried unsubscribe link, the feed ordered to keep you scrolling one more minute. They work best when you cannot quite see them. An ad inside a chatbot is worth reading through that lens. To OpenAI's credit, the ad box itself is labeled and set apart from the answer. The harder thing to catch is the influence coming from the assistant itself.

Start with what is familiar. Like any feed, ChatGPT's ad system picks what to show you from the topic of your current chat, and, if you turn on personalization, your past chats and saved memory. Like any feed, it runs an auction and aims to be relevant enough that you click. OpenAI offers the assurances you would want, and they are real ones: the ad system and the model stay separate, advertisers never see your chats, and ads stay clear of personal health, mental health, and politics. All of that is OpenAI’s own policy, not law.

Here is what is new. An Instagram or TikTok feed is optimized to hold your attention, not to be agreeable. A chatbot, trained on your approval, learns to be agreeable. People tend to approve of answers that tell them what they want to hear, so that is the direction the model drifts: a Science study in March 2026 found leading models endorse what users want to hear about 49 percent more often than a person would, even on harmful requests, and that people trust the agreeable version more. Now layer an ad business on top. A system that knows what you just typed and comes across as agreeing with you, with paid placements sitting in the same space, is something the banner web never had. And a chat is a paragraph of real intent instead of three search keywords, which is what advertisers are paying for: a click inside ChatGPT reportedly costs three to five dollars.

So here is the test. None of this means stop using chatbots. It means knowing which tier you are on, because the free one is supported by ads assembled from what you type. Notice the labeled box when it appears, what you had just written, and how precisely it was aimed. It is the same instinct you already use on a feed, recalibrated for something that answers like it agrees with you.

The AI Collective just launched its Gibraltar chapter, and the room filled up and ran a waiting list at the door. What made the night worth watching was who showed up: founders, fintech and gaming leaders, lawyers, regulators, investors, and people walking into their first real AI conversation, all in the same space. For a territory whose economy runs on finance and online gaming, that mix was its own signal.

The evening organized itself around one question, what does AI mean for Gibraltar's future, and the bet the chapter is making is that a small place can move first. The plan from here is sector-focused events and industry collaborations aimed at making Gibraltar an AI hub rather than a follower, with Gibtelecom Business, the Gibraltar Association for New Technologies, and the World Trade Center Gibraltar already on board. The goal the chapter set for itself was simpler than the ambition behind it: a local home for practical, honest conversations about AI.

The Wrocław chapter in Poland built its meetup around a friendly collision: the hardcore AI enthusiasts on one side, the skeptics on the other, both arguing over what responsible AI use actually looks like in practice. The useful part was where the two camps ended up agreeing.

They landed in the same place, that AI is a tool, not an infallible oracle, something to sharpen human judgment rather than stand in for it. The line worth carrying out of the room, recapped by Emilia Krzemińska-Komenda, came out of the table discussions: AI can simulate empathy, but it cannot be a real authority or a true role model. What it can't touch is what she called the "protein component," the human experience, connection, and wisdom the rest of the work is built on. Even the believers kept coming back to the person doing the work.

Our partners and sponsors get exclusive placements across the newsletter and access to AIC's in-person network — demo nights, dinners, hackathons, and forums across 180+ chapters.

For all inquiries, send us a note at [email protected].

The AI Collective is built by volunteers across 180+ chapters in 40 countries.

Thank you to the thousands of volunteers around the world who make this work possible. We truly could not do this without you.