In October 2024, Sam Altman said advertising was "like a last resort for us as a business model." This week, OpenAI announced it's introducing ads to the free and Go tiers of ChatGPT.
That's a fast trip to the last resort.
Let's be clear about what this means: OpenAI has raised billions in funding at a $157 billion valuation. It charges $20/month for Plus subscriptions. It has enterprise contracts with major corporations. And it still needs advertising revenue to make the economics work.
That should tell you everything about AI unit economics.
The company is positioning this as expanding access - "advertising allows us to serve more users for free." That's technically true. It's also conveniently profitable for OpenAI. Advertising at scale is lucrative. Google makes $200+ billion annually from ads. If even a fraction of ChatGPT's hundreds of millions of users see ads, that's real money.
The announcement specifies ads will appear initially in the free and Go tiers. Initially. That word is doing a lot of work. If you believe premium subscribers are permanently safe from ads, I have a Netflix ad-free tier to sell you from 2015.
The progression is predictable: Start with free users (they can't complain, they're not paying). Then add ads to lower-paid tiers (they're paying less, so ads offset costs). Eventually introduce a new "Premium Plus Ultra" tier that's actually ad-free while the regular Plus tier gets ads (but hey, they're "tastefully integrated").
One Reddit user in the source thread made an excellent point: "No thank you, I don't want product recommendations in my answers when I make important health emergency related questions."
