OpenAI Launches ChatGPT Ads Platform, Says Paying Plus/Pro/Business Users Won't See Them
- OpenAI opened ads.openai.com, a self-serve Ads Manager letting advertisers create campaigns, upload ad details in bulk, and track impressions, clicks, and conversions inside ChatGPT.
- Early advertisers named in the launch include Best Buy, Lowe's, and VistaPrint; Best Buy's VP of Media, Amy Adams, said the company is "encouraged by the early results."
- OpenAI states ads are shown only to free-tier users: "We do not show ads to users on Plus, Pro, or any Business plan, or to any users who tell us or we predict are under the age of 18."
- The move comes after Sam Altman previously called advertising "a last resort for us for a business model" at a Harvard talk in 2024, and follows reports that OpenAI's IPO has slipped to 2027.
- OpenAI says ads will be "clearly labeled" and kept "separate from answers," with users controlling how their data is used, though several Hacker News commenters reported receiving ads unrelated to their chat topic.
Hacker News 의견들
Contrast this ad launch with Altman's own words: he called ads 'a last resort for us for a business model' back at Harvard 2024. Guess that resort just got a lot closer.
IPO has been delayed to 2027, so this doesn't look like a sign of strong financial health if they're pushing ads now.
Timing is bold given the Open Models vs Proprietary debate. Feels like a really easy decision when framed as chasing free-tier revenue.
The median non-technical ChatGPT user doesn't pay, won't self-host an open model, and doesn't use an adblocker, so ads make total sense for that crowd.
People care about the pro/paid tiers getting ads eventually. Of course they'll say Plus and Pro stay ad free now, just like Netflix and Prime once said the same thing.
OpenAI's own help doc says: 'We do not show ads to users on Plus, Pro, or any Business plan, or to any users who tell us or we predict are under the age of 18.' So free tier only, for now.
I keep getting ads that make zero sense next to my chat, like talking hardware and getting a wood ad. So much for 'relevant ads that fit naturally into the experience'.
This is way better than baking ads directly into the ChatGPT answers themselves. Standard sidebar ads I can at least recognize and ignore. Though I bet they'll eventually do both.
Google Ads started unobtrusive too and we all know where that ended up. Let's see if Anthropic keeps a straight face and follows OpenAI down this road.
'Clearly labeled' and 'separate from answers' is the same rock solid trust promise every platform makes right before it slowly turns into an ad-soaked mess over a few years.