~20,000 founders and comms leaders start their week with The Brief.
Join them to get insights into the media, communications, and PR industry.

Last month, we talked about AI buying the media. But r-r-r-remix: now brands are buying ad space inside AI.

Remember when we were all freaking out about AI writing emails? lol suckers.

ChatGPT Ads launched this past February, and OpenAI is building a full Ads Manager like it binge-watched a docuseries about Google. 

Ads are now running inside ChatGPT conversations, and buy-in starts at a cool $200,000. Which sounds pretty steep, but somewhere, a CEO just asked, "What's our ChatGPT strategy?" and a budget magically appeared.

Plus, an early Microsoft study found that “AI users on Copilot engaged far more often with ads” – a staggering 73% higher click-through rate. Damn.

The ads pop up at the end of responses/answers (like this ad we got when we recruited ChatGPT to help compare potential new couch dimensions):

OpenAI shared its ad principles, promising that ads won’t influence the answers ChatGPT produces and that conversations stay private. The info won’t be shared with brands. Very noble, indeed!

Two red flags (read: PR problems). 

🚩 #1: People are going to be apprehensive about the principles

We're probably not alone in being a little skeptical (you can’t blame us for being suspicious of the world right now!). OpenAI's principles sound great on paper…but we've all seen it before. 

Platform promises user-first values, platform finds a loophole, and suddenly every answer starts to feel a little suspect.  👀

If users can't trust that ChatGPT's recommendations are truly independent from the ads running alongside them, the whole value proposition unravels. 

So would the ad spend behind it. 🫤 

🚩 #2: People perceive LLMs as objective authorities

Not ad platforms.

When sponsored mess starts clogging up conversations and piggybacks on every response, what you’re getting back doesn’t seem unbiased anymore. And it’ll be on PR to handle that authenticity backlash. 

Ads with ChatGPT can’t be controlled. 💃 (but, same)

Businesses have no say in which conversations the ads will appear in. 

In Google, you bid on specific keywords. Here, not so much. 

Imagine agreeing to sponsor a conversation before knowing whether it's about productivity hacks, home improvement, or someone's full-blown existential crisis.

We’ve got a whole new crisis management situation. 

In one sense, it’s the usual job – anticipate risk beforehand. PR folks will now have to consider how ads will show up in different conversations. What kind of responses would those get?

Example: if a weight-loss brand bought an ad on ChatGPT, it could appear when someone searches for ways to eat healthy and lose weight. 

But it could also pop up when someone is looking for coping strategies for anxiety or other mental health concerns.

Very different vibes. 🫤

We all know AI occasionally decides facts are more of a "creative suggestion."

Sometimes false information is generated, like incorrect product features, prices, or straight-up incorrect facts. 

If your ad gets served with something that leads someone astray, how would that reflect on you?

Now, it’s true. We haven’t come across any huge headline about this. But it’s the kind of thing that only needs to happen once. And no one wants it to be them. 

Search? Social? Comms? Companies need to decide whether it lives in-house (and if so, where) or if you’re gonna go third-party. 

Either way, you need some stellar cross-functional alignment.

ChatGPT ads straddle this weird space between paid media, brand rep, and messaging/strategy. That covers multiple teams. 

Without alignment, the search team could buy one message, comms is pushing another, brand is somewhere else entirely, and congrats, you've accidentally created your own PR multiverse.

Sorry, but we’d like to avoid crisis management of self-inflicted crises. 

Don’t read this and go all buck wild, throwing your eggs in one basket. 

Brands that rely solely on paid placement risk losing organic authority, and that’s necessary to be cited by AI. 

Paid gets you in the room.

But earned keeps you in the answers.

We don’t really know what the standard business practice will be yet. shrug

Everyone’s basically beta testing how this whole thing is gonna work.

But we’ll keep you posted.

If you need us, we'll be over here conducting extremely scientific research by asking ChatGPT increasingly unhinged questions and seeing which ads show up. 💁‍♀️

Ash & Lizzy

Who's going where and why it matters. Not just job shifts - power dynamics, layoffs, and who's headed out.

🤖 If you still weren’t fully bought into the conversation around AI, it might be time to change your tune. Yahoo Finance announced a dedicated hub (in partnership with Axios, Fortune Media, TechCrunch, and Fast Company) just for AI news and analysis.

🧳 Corbin Bolies is now at Variety, covering the intersection of AI and entertainment.

🧳 Ken Brown is now CNBC’s managing editor of digital and editorial strategy.

🧳 Dakin Campbell has joined The Information to follow the money behind the AI boom.

🧳 Grace Kay has moved from Business Insider to The Information, where she will continue to cover Elon Musk’s empire (a fascinating beat, indeed!).

👀 New York Times employees are gearing up for a fight about how the news organization is using AI tools to track and evaluate employee performance and activity.

🧳 Webb Wright has joined Gizmodo as a senior reporter covering AI.

Get to know Ashley and Lizzy, learn more about The Colab, or pass The Brief along to a friend who should be reading this too.