OpenAI Is Becoming a Shopping Mall, and I'm Not Sure How to Feel About It

But Instacart checkout? Target shopping? A self-serve ads platform? This is a different kind of company.

I should be clear: I'm not saying this is evil or even necessarily bad. Lots of good companies make money from commerce and advertising. Google does. Meta does. Amazon obviously does. But those companies never claimed to be nonprofit research labs focused on ensuring AI benefits all of humanity. OpenAI did.

What We Don't Know

There's a lot that remains unclear here, and I think that's worth acknowledging.

First, we don't know how much of ChatGPT's development resources are now going toward commerce features versus, you know, making the AI smarter or safer. OpenAI doesn't break that out publicly.

Second, the advertising privacy claims are vague. They say conversations are "separate" from ads, but ChatGPT is fundamentally a conversation product. If I ask about running shoes and then see running shoe ads, is that separation? The company didn't disclose exact figures on how targeting will work.

Third, I couldn't find clear information on revenue splits with partners. Is Instacart paying OpenAI for placement? Is OpenAI taking a cut of transactions? Both? The announcements are light on these details.

What I do know is that OpenAI is moving fast. The ads beta launched with measurement tools and CPC bidding already built. That's not a test, that's infrastructure. You don't build infrastructure for something you're not committed to.

The Bigger Picture

Here's where I land on this, tbh.

OpenAI has every right to build a sustainable business. The alternative is running out of money or becoming entirely dependent on Microsoft, neither of which seems great for their stated mission. And making ChatGPT useful for everyday tasks like shopping isn't inherently bad. It might actually be good! People need groceries.

But I think we should be honest about what's happening. The company that Sam Altman describes as building AGI (artificial general intelligence, the kind that could theoretically surpass human capabilities) is now also in the business of helping you buy paper towels and showing you ads.

Those two things can coexist. But they create different incentives. A research lab optimizes for capability and safety. A commerce platform optimizes for engagement and transactions. An ad-supported product optimizes for attention. These aren't the same goals, and pretending they are feels naive.

I'm genuinely uncertain whether this matters in the long run. Maybe OpenAI will use commerce revenue to fund more safety research. Maybe the shopping features will bring AI to people who wouldn't otherwise use it. Maybe I'm overthinking a grocery delivery integration.

But I keep coming back to that original pitch: a capped-profit company, founded to ensure AI benefits humanity, backed by billions in nonprofit research. That company is now running a self-serve ads platform and processing Target checkouts.

Something changed. I'm just not sure anyone at OpenAI wants to talk about what.

Sarah Williams · 51 mins ago · 7 min