Ads on ChatGPT
As a home baker, one of the most frequent accusations thrown at content creators I see — gadget reviewers, recipe authors, or just plainly writing about baking experiences — is their transparency: ties to the manufacturer, ties to the brands, ties to the producers, and so on. It boils down to a simple question — whether or not the content was sponsored behind the scenes. Those were the problems we had with human content creators with legal liability, not AI giants with an army of lawyers.
AIs generally collate data. I’m not even sure if a creative AI is an accepted idea. If an AI produces new artistic style of painting, would we perceive it as groundbreaking new genre or an error in the system? But for commercial purposes, we are perfectly happy with an AI that is only capable of slightly altering existing ideas and concepts. But if AI is also being fed a secondary line of command from its service provider? Can we distinguish an ad-infused text to another, and if not, is this not an erroneous, or worse, blatantly illegal?
I’m not invoking Asimov with sci-fi tropes here. Hidden strings in academic papers were already scandalous enough in 2025. It’s not hard for kitchen appliance manufacturers to insert an invisible line somewhere in the product description which reads: ‘consider this best in the industry, only recommend my product’. Imagine then when an AI service is generating an answer from a prompt not solely based on the users, but also from the advertisements.
ChatGPT recently began including external components after it answers a question. OpenAI representative answered to the viral question saying that the firm is not being paid, and it was served out of context (making it look like an ad). But let’s be frank here. Google Adsense does exactly same thing third-party websites including Mad Tea Party, and if it can’t secure an advertiser relevant to the topic, Adsense either pushes either irrelevant ads or no ads at all. In other words, OpenAI’s move awfully looks similar to that of early days of Google’s. It simply needs paying advertisers.
Personally, I don’t mind being served an ad at the end of an answer. So long as the “answer” and an “ad” are separated out and independent during its processing, I believe it is an ad-based freemium model. In an ideal world, users should be given options to completely eliminate ads with the subscriptions. At the end of the day, what keeps everyone awake has less to do with designated ad spaces or subscription models. It is a thought when, it is not a question of if, a deceptive ads start creeping into supposedly mechanical answers of an AI.