If AI follows the same path as social networks, answers will start to depend on who pays to push their message. And if you do not want paid answers, you will have to pay yourself.
Here in the United States, OpenAI announced that in the coming weeks it will start testing ads in ChatGPT. The test involves adult users on the free version and on the Go plan. Plus, Pro, and Enterprise stay ad-free. For now, ads should appear in a separate section at the bottom of answers, clearly labeled.
This is exactly how social networks started. Then it ended with scammers buying scam ads, often without being stopped. In many cases, up to 10% of revenue comes from fraudulent advertising.
OpenAI also says it wants to protect trust and privacy: no ads for users under 18, no ads next to topics like politics or health, and no selling conversation data to advertisers.
The real issue is not the banner itself. The issue is that when an assistant becomes the place where we ask what to buy, which bank to choose, or which doctor to trust, the order of answers becomes power. Today ads are separate and visible. Tomorrow they could be mixed into the text, a “sponsored” suggestion that looks like neutral advice. Visible ads are easy to spot. Ads blended into answers are not.
Think about everyday family questions. “What is the best car insurance in New York for a new driver?” “Which bank account is best if I have a salary and a mortgage?” “Which app should I use to invest small amounts safely?” “Which English course should I choose for my child?”
There is no single truth in these questions. There are rankings, based on criteria. If someone pays to rank higher, the “useful truth” changes. No lies are needed. You just highlight one feature, ignore another, and choose one example instead of others.
OpenAI says answers are not influenced by ads. That is good, but influence is not only in the final sentence. It is also in what is shown, what is left out, and how priorities are built. This is the same mechanism that changed web search over time.
What should we do? Use AI as a tool, not as a judge. For money, contracts, or health, always double-check with independent sources. When advice looks perfect, ask what is missing: alternatives, real costs, limits, negative reviews, conditions.
Otherwise, we risk becoming puppets guided by whoever pays to shape our opinions.
#ArtificialDecisions #MCC